Dataset Preview
The full dataset viewer is not available (click to read why). Only showing a preview of the rows.
The dataset generation failed because of a cast error
Error code: DatasetGenerationCastError
Exception: DatasetGenerationCastError
Message: An error occurred while generating the dataset
All the data files must have the same columns, but at some point there are 2 new columns ({'page', 'content'}) and 2 missing columns ({'instruction', 'output'}).
This happened while the json dataset builder was generating data using
hf://datasets/ResourceFX/Sanskrit/sanskrit book json data.json (at revision 51061f60d4dd72ba612635acbc4ce806fda5dab1)
Please either edit the data files to have matching columns, or separate them into different configurations (see docs at https://hf.co/docs/hub/datasets-manual-configuration#multiple-configurations)
Traceback: Traceback (most recent call last):
File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/builder.py", line 1870, in _prepare_split_single
writer.write_table(table)
File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/arrow_writer.py", line 622, in write_table
pa_table = table_cast(pa_table, self._schema)
File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/table.py", line 2292, in table_cast
return cast_table_to_schema(table, schema)
File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/table.py", line 2240, in cast_table_to_schema
raise CastError(
datasets.table.CastError: Couldn't cast
page: int64
content: string
-- schema metadata --
pandas: '{"index_columns": [], "column_indexes": [], "columns": [{"name":' + 312
to
{'instruction': Value(dtype='string', id=None), 'output': Value(dtype='string', id=None)}
because column names don't match
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 1420, in compute_config_parquet_and_info_response
parquet_operations = convert_to_parquet(builder)
File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 1052, in convert_to_parquet
builder.download_and_prepare(
File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/builder.py", line 924, in download_and_prepare
self._download_and_prepare(
File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/builder.py", line 1000, in _download_and_prepare
self._prepare_split(split_generator, **prepare_split_kwargs)
File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/builder.py", line 1741, in _prepare_split
for job_id, done, content in self._prepare_split_single(
File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/builder.py", line 1872, in _prepare_split_single
raise DatasetGenerationCastError.from_cast_error(
datasets.exceptions.DatasetGenerationCastError: An error occurred while generating the dataset
All the data files must have the same columns, but at some point there are 2 new columns ({'page', 'content'}) and 2 missing columns ({'instruction', 'output'}).
This happened while the json dataset builder was generating data using
hf://datasets/ResourceFX/Sanskrit/sanskrit book json data.json (at revision 51061f60d4dd72ba612635acbc4ce806fda5dab1)
Please either edit the data files to have matching columns, or separate them into different configurations (see docs at https://hf.co/docs/hub/datasets-manual-configuration#multiple-configurations)Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
instruction
string | output
string |
|---|---|
Summarize page 1: | BIBLIOTHEK
INPOGERMAMSCHER GRAMMATIKEN
BEARBEITET VON
F.BUCHELER, H,HUBSOHMANN,A,LESKIEN, G,MEYER,
E.SIEVERS,H,WEBER, W,D,WHITNEY,E,WINDISOH,
BAND II.
ASANSKRIT GRAMMAR, INCLUDING BOTH THE CLASSICAL LANGUAGE, AND
THEOLDER DIALECTS, OFVEDA ANDBRAHMANA
BYWILLIAM DWIGHT WHITNEY.
LEIPZIG,
DRUCK UNDVERLAG VONBREITKOPF UNDHARTEL.
1879.
|
Summarize page 2: | PREFACE.
Itwas inJune, 1875. asIchanced tobeforaday or
two inLeipzig,that Iwas unexpectedlyinvited toprepare
theSanskrit grammarfortheIndo-Europeanseriesprojected
byMessrs. Breitkopf andHartel. After some consideration,
andconsultation with friends,Iacceptedthetask, andhave
since devoted toitwhat time could bespared fromregular
duties,after thesatisfaction ofengagementsearlier formed.
Ifthedelay seems along one,itwasnevertheless unavoid-
able;and Iwouldgladly,intheinterest ofthework itself
havemade itstilllonger.Inevery suchcase,itisnecess-
arytomake acompromise between measurably satisfyinga
present pressing need, anddoing thesubjectfullerjustice
atthecostofmore time; and itseemed asifthecall for
aSanskrit grammar onasomewhat differentplanfrom those
alreadyinuse excellent assome ofthese inmany respects
are- -was urgent enoughtorecommend aspeedy com-
pletionofthework begun.
Theobjects hadespeciallyinview inthepreparation
ofthisgrammar have been thefollowing:
Tomake apresentationofthe facts ofthelanguage
primarilyastheyshow themselves inuseintheliterature,
andonly secondarilyasthey arelaiddown bythenative
grammarians. The earliest European grammars werebythe
necessityofthecasechiefly founded ontheir nativeprede-
cessors;andatraditional method wasthus established which
hasbeen perhaps somewhat tooclosely adheredto,atthe
expenseofclearness andofproportion,aswell asofscien-
tific truth. Accordingly, myattention hasnotbeen directed
toward aprofounder studyofthegrammaticalscience ofthe
Hindu schools :their teachingsIhave been contented totake
|
Summarize page 3: | viPREFACE.
asalready reportedtoWestern learners intheexisting
Western grammars.
Toinclude also inthepresentationtheforms andcon-
structions oftheolder language,asexhibited intheVeda
andtheBrahmana. Grassmann's excellent Index-Vocabulary
totheRig-Veda, andmyownmanuscript onetotheAtharva-
Veda (whichIhopesoon tobeable tomakepublic), gave
meinfull detail thegreatmass ofVedic material;andthis,
withsome assistance frompupilsandfriends,Ihavesong]it
tocomplete,asfarasthecircumstancespermitted, from the
other Vedic textsandfrom thevarious works oftheBrah-
manaperiod,bothprintedandmanuscript.
Totreat thelanguage throughout asanaccentedone,
omitting nothingofwhat isknownrespecting thenature of
theSanskrit accent,itschangesincombination andinflection,
andthetone ofindividual words- -being, inallthis, ne-
cessarily dependent especially upon thematerialpresented
bytheolder accentuated texts.
Tocast allstatements; classifications, andsoon,intoa
form consistent with theteachiogs oflinguistic science. In
doing this,ithasbeennecessarytodiscard afew ofthe
long-used andfamiliar divisions andterms ofSanskrit gram-
mar--forexample, the classification andnomenclature of
"specialtenses" and"general tenses" (whichissoindefen-
sible thatonecanonlywonder atitshaving maintained itself
solong),theorderandterminologyoftheconjugation-classes,
theseparation intreatment ofthefacts ofinternal andex-
ternal euphonic combination, andthelike. Butcarehasbeen
taken tofacilitate the transition from theoldtothenew;
andthechanges,itisbelieved, willcommend themselves
tounqualified acceptance. Ithasbeensought also tohelp
anappreciation ofthecharacter ofthelanguage byputting
itsfacts asfaraspossible intoastatistical form. Inthis
respect thenative grammar isespecially deficient andmis-
leading.
Regard hasbeenconstantly had tothepractical needs
ofthelearner ofthelanguage, and ithasbeenattempted,
byduearrangement andbytheuseofdifferent sizes of
|
Summarize page 4: | PREFACE. vii
type,tomake thework asusable byonewhoseobject
itistoacquireaknowledgeoftheclassical Sanskrit alone
asthose are inwhich the earlier forms arenotincluded.
Thecustom oftransliteratingallSanskrit words intoEuro-
pean characters, which hasbecome usual inEuropean Sans-
kritgrammars, is,asamatter ofcourse, retained through-
out; and. because ofthedifficultyofsetting even asmall
Sanskrit typewithanythingbutalarge European,itis
practicedalone inthesmaller sizes.
While thetreatment ofthe facts ofthelanguagehas
thusbeenmade ahistorical one, within the limits ofthe
language itself,Ihave notventured tomake itcomparative,
bybringingintheanalogous forms andprocessesofother
related languages. Todothis,inaddition toallthatwas
attempted beside, would have extended thework, both in
content and intime ofpreparation,farbeyondthelimits
assignedtoit.And, having decided toleave outthis ele-
ment, Ihave done soconsistently throughout. Explanations
oftheoriginofforms have alsobeen avoided, forthesame
reason and forothers, which hardlycall forstatement.
Agrammarisnecessarilyingreat partfounded onits
predecessors, and itwould beinvain toattempt anacknowl-
edgmentindetail ofalltheaidreceived from other schol-
ars. Ihavehad athand always especiallytheveryschol-
arlyand reliable briefsummaryofKielhorn, the fulland
excellent work ofMonier Williams, thesmaller grammarof
Bopp (awonder oflearning andmethod forthetimewhen
itAvasprepared ,andthevolumes ofBenfey and Mtiller.
Asregards thematerial ofthelanguage, nootheraid,of
course, hasbeen atallcomparable with thegreatPeters-
burg lexicon ofBohtlingk andRoth, theexistence ofwhich
gives byitself anew character toallinvestigationsofthe
Sanskritlanguage. What Ihave notfound there orinthe
specialcollections made bymyselforbyothers forme,I
have called below "notquotable*'-aprovisional designa-
tion^ necessarilyliable tocorrection indetail bytheresults
offurther researches. Forwhat concerns theverb,itsforms
and their classification anduses,Ihave had, asevery one
|
Summarize page 5: | viiiPREFACE .
must have, byfarthemost aidfromDelbruck. inhisAlt-
indisches Verb urnand hisvarious syntacticalcontribu-
tions. Former pupilsofmyown. Prof. Avery and Dr.
Edgren. have alsohelpedme. inconnection with thissub-
jectandwith others, inawayandmeasure that calls for
public acknowledgment.Inrespecttotheimportant matter
ofthedeclension intheearliest language.Ihavemade great
useoftheelaborate paperintheJourn. Am. Or.Soc.(print-
ingcontemporaneouslywith thiswork, andused byme
almost, butnotquite,totheend ofthesubject) bymy
formerpupilProf.Lanman; mytreatment ofitisfounded
onhis.Mymanifold obligationstomyown teacher. Prof.
Weber ofBerlin, also requiretobementioned :among other
things,Iowe tohim theuseofhiscopiesofcertain un-
publishedtexts oftheBrahmanaperiod,nototherwise access-
ible tome;andhewaskindenoughtolookthrough with
memywork initsinchoate condition, favoring mewith
valuablesuggestions. For this lastfavor Ihave likewise to
thank Prof.Delbruck who, moreover, hastaken thetrouble
toglance over foralikepurposethegreater part ofthe
proof-sheets ofthegrammar, astheycame from thepress.
ToDr.L.Schroder isduewhatever useIhave been able
tomake(unfortunately avery imperfect one)oftheimport-
antMatriayani-Sanhita.
Ofthedeficiencies ofmywork Iam. Ithink, notless
fullyaware thananycritic ofit.even theseverest,islikely
tobe.Should itbefound toanswer itsintended purpose
wellenoughtocome toanother edition, myendeavor will
betoimprove andcomplete it;and Ishall begratefulfor
any corrections orsuggestions which mayaidmeinmak-
ingitamore efficienthelptothestudyoftheSanskrit
language and literature.
GOTH A,July1879.
W.D.W.
|
Summarize page 6: | INTRODUCTION.
BRIEF ACCOUNT OFTHEINDIAN LITERATURE.
Itseems desirable togivehere such asketch ofthe
historyofIndian literature asshall show the relation to
oneanother ofthe differentperiods andforms ofthelan-
guagetreated inthefollowing grammar,and theposition
oftheworks therequoted.
Thename "Sanskrit" (samskrta, 1087 d,'adorned, elab-
orated,perfected'), which ispopularly appliedtothewhole
ancient andsacredlanguageofIndia, belongs moreproperly
onlytothat dialect which, regulatedandestablished bythe
labors ofthenativegrammarians,hasledforthe lasttwo
thousandyearsormore anartificiallife,likethat ofthe
Latinduringmost ofthesameperiodinEurope,asthe
written andspoken means ofcommunication ofthelearned
andpriestlycaste
;andwhich even atthepresent dayfills
that office. Itisthusdistinguished,ontheonehand, from
the later and derived dialects asthePrakrit, forms of
language which have datable monuments from asearlyas
thethirdcenturybefore Christ, andwhich arerepresented
byinscriptions and coins, bythespeechoftheuneducated
characters intheSanskrit dramas(see below), andbya
limited literature;thePali, aPrakritic dialect which became
thesacredlanguageofBuddhism inFarther India, and is
|
Summarize page 7: | x INTRODUCTION.
still inservice there assuch;andyetlaterandmore altered
tongues formingthetransition tothelanguagesofModern
India. And, ontheother hand, itisdistinguished,but
verymuch lesssharplyandwidely,from theolder dialects
orforms ofspeech presentedinthecanonical literature,
theVeda andBrahmana.
Thisfact,ofthefixation bylearned treatment ofan
authorized mode ofexpression,which should thenceforth be
usedaccordingtorule intheintercourse oftheeducated,
isthecardinal oneinIndianlinguistic history; and asthe
nativegrammaticalliterature hasdetermined theform of
thelanguage,soithas also toalargeextent determined
thegrammaticaltreatment ofthelanguage byEuropean
scholars.
Much inthehistoryofthelearned movement isstill
obscure, andopinionsareatvariance even astopointsof
prime consequence. Onlytheconcluding works inthedevel-
opmentofthegrammatical science havebeenpreservedto
us;andthough theyareevidentlytheperfectedfruits ofa
longseries oflearnedlabors, therecords ofthelatter are
lostbeyond recovery. The timeandtheplaceofthecre-
ation ofSanskrit areunknown;and astoitsoccasion, we
haveonlyourinferences andconjecturestorelyupon.It
seems, however, altogether likelythatthegrammaticalsense
oftheancient Hindus wasawakened ingreat measure by
theirstudyofthetraditional sacred texts, andbytheircom-
parison ofitsdifferentlanguagewith that ofcontemporary
use. Itiscertain that thegrammatical studyofthose texts
(gakhas, lit'ly 'branches'), phonetic and other, waszealously
andeffectively followed intheBrahmanic schools;this is
attestedbyourpossession ofanumber ofphonetico-gram-
maticaltreatises, firatigaJchyas (prati $ahham, 'belongingto
each several text1
),onehavingforsubject eachprincipal
Vedictext,andnotingallitspeculiarities ofform; these,
bothbythedepth and exactness oftheirown researches
andbythenumber ofauthorities whichthey quote, speak
plainlyofalivelyscientificactivity continuedduringalong
time. Whatpart, ontheother hand, thenotice ofdiffer-
|
Summarize page 8: | xii'INTRODUCTION.
limits oftheartificialityofthisprocessisnotyetknown.
The attention ofspecialstudents oftheHindu grammar
andthesubjectissointricate and difficult that thenumber
isexceedinglysmall ofthosewhohave mastered itsuffi-
cientlytohave acompetent opinion onsuchgeneral matters)
hasbeen hitherto mainlydirected toward determining what
theSanskrit accordingtoPaninireallyis,towardexplaining
thelanguagefrom thegrammar. Arid, naturally enough,
inIndia,orwherever elsetheleading objectistolearn to
speakandwrite thelanguage correctlythatis,asauthor-
izedbythegrammariansthat isthepropercourse to
pursue. This, however,isnottheway reallytounderstand
thelanguage. The timemust soon come, orithascome
already, when theendeavor shall beinstead toexplainthe
grammarfrom thelanguage;totest inalldetails, sofar
asshall befoundpossible,thereason ofPanini' srules
(whichcontain notalittle thatseemsproblematical,oreven
sometimesperverse);todetermine what andhowmuch
genuine usage hehadeverywhereasfoundation, andwhat
traces maybeleftintheliterature ofusages possessingan
inherentlyauthorized character, thoughunratified byhim.
Bytheterm "classical'' or"later"language, then, as
constantlyused below inthegrammar,ismeant thelan-
guageofthoseliterary monuments which arewritten incon-
formitywith therules ofthenative grammar:virtually,the
wholeproperSanskrit literature. Foralthough partsofthis
aredoubtless earlier than Panini,itisimpossibletotell
justwhatparts,orhow fartheyhaveescapedintheirstyle
thelevelling influence ofthegrammar. The whole, too,
maybecalled sofaranartificial literature asitiswritten
inaphonetic form(seegrammar, 103)which never can
havebeen atrulyvernacular andlivingone.Nearlyallof
itismetrical :notpoetic worksonly,butnarratives, histories
(sofarasanything deservingthatname canbesaid toexist),
and scientific treatises ofevery variety,aredone intoverse;
aprose andaprose literature(exceptinthecommentaries)
hardlyhasanexistence. Oflinguistic historythere isnext
tonothinginitall;butonlyahistoryofstyle,and this
|
Summarize page 9: | INTRODUCTION. xiii
forthemostpartshowingagradual depravation, anincrease
ofartificiality andanintensification ofcertain more unde-
sirable features ofthelanguage such astheuseofpas-
sive constructions andofparticiplesinstead ofverbs, and
thesubstitution ofcompoundsforsentences.
Thisbeingthecondition ofthelaterliterature,itisof
somuch thehigher consequencethat there isanearlier
literature,towhich thesuspicionofartificialitydoes not
attach, orattaches atleastonlyinaminimaldegree, which
hasatrulyvernacular character, andabounds inproseas
well asverse.
The results oftheveryearliestliterary productiveness
oftheIndianpeoplearethehymnswith which, whenthey
hadonlycrossed thethreshold ofthecountry, andwhen
theirgeographicalhorizon was still limited tothe river-
basin oftheIndus with itstributaries, they praisedtheir
gods,the deified powersofnature, andaccompanied the
rites oftheircomparatively simple worship. Atwhatperiod
these weremade andsungcannot bedetermined withany
approachtoaccuracy:itmayhavebeen asearlyas2000
B.C.They were longhanded down byoraltradition, pre-
served bythe care, andincreased bytheadditions and
imitations, ofsucceeding generations;themass wasever
growing, and, with thechangeofhabits and beliefs and
religious practices, wasbecoming variously applied sung
inchosen extracts, mixed with other material intoliturgies,
adapted withmore orlessofdistortion tohelptheneeds
ofaceremonial which wascomingtobeofimmense elab-
oration andintricacy. And, atsome time inthecourse
ofthishistory,there wasmade forpreservationagreatcol-
lection ofthehymn-material, mainlyitsoldest andmost
genuine part,totheextent ofover athousand hymns and
tenthousandverses, arranged accordingtotraditional author-
shipand tosubject andlengthofhymn:this collection is
theRig-Veda, -Veda ofverses(re)orhymns'. Other col-
lections weremade also outofthesamegeneralmass of
traditional material :doubtlesslater, althoughthe inter-
relations ofthisperiodareasyettoounclear toallow of
|
Summarize page 10: | xiv INTRODUCTION.
ourspeakingwith entire confidence astoanythingconcern-
ingthem. Thus, theSama- Veda. 'Veda ofchants(saman}\
containing onlyabout asixth asmuch, itsversesnearlyall
found intheRig-Vedaalso, butappearingherewithnume-
rous differences ofreading;these werepassages puttogether
forchantingatthesoma-sacrifices. Again,collections called
bythecomprehensivename ofYajur-Veda, 'Veda ofsac-
rificial formulas (yajusV:these contained notverses alone,
butalsonumerous prose utterances, mingledwith theformer,
intheorder inwhichtheywerepractically employedin
theceremonies; theywerestrictly liturgicalcollections. Of
these, there areinexistence several texts, which have their
mutual differences: theVajasaneyi-Samhita (intwoslightly
discordant versions, Madhyandina andKanvd,sometimes
also called theWhiteYajur-Veda;and thevarious and
considerably differingtexts oftheBlack Yajur-Veda. namely
theTaittirlya-Samhita,theMaitrayam-Samhita,and the
Kathaka (thetwo lastnotyetpublished). Finally,another
historical collection, liketheRig-Veda,butmade upmainly
oflaterand lessacceptedmaterial, andcalled (amongother
less current names) theAtharva-Veda, 'Veda oftheAth-
arvans(alegendary priestly family)';itissomewhat more
than half asbulkyastheRig-Veda, andcontains acertain
amount ofmaterialcorrespondingtothat ofthe latter, and
also anumber ofbriefprose passages. Tothis last col-
lection isvery generallyrefused intheorthodox literature
thename ofVeda; butforusitisthemostinterestingof
all. after theRig-Veda, because itcontains thelargest
amount ofhymn-material (ormantra, asitiscalled, in
distinction from theprose brahmana], and inalanguage
which, though distinctlylessantique than that oftheother,
isneverthelesstrulyVedic. Two versions ofitareextant,
oneofthemonlyinasingle knownmanuscript.Anotinsignificant bodyoflikematerial, andofvarious
period (although doubtless inthemainbelongingtothe
latest time ofVedicproductiveness, and inpart perhaps
theimitative work ofayetmoremoderntime),isscattered
through thetexts tobelater described, theBrahmanas and
|
Summarize page 11: | INTRODUCTION. xv
theSutras. Toassemble and siftandcompareitisnow
oneofthepressingneeds ofVedicstudy.
Thefundamental divisions oftheVedic literature here
mentioned allhave had their various schools ofsectaries,
each ofthese with atext ofitsown. showing some differ-
ences from those oftheother schools :butthose mentioned
above are allthat arenowknown tobeinexistence; and
thechance ofthediscoveryofothers grows every year
smaller.
The labor oftheschools intheconservation oftheir
sacred textswasextraordinary, andhasbeen crowned with
such success that thetext ofeach school, whatever may
be itsdifferences from those ofother schools, isvirtually
without variousreadings, preservedwith allitspeculiarities
ofdialect, and itssmallest andmostexceptionaltraits of
phonetic form, pure andunobscured. Itisnottheplace
here todescribe themeans bywhich, inaddition tothe
religiouscare ofthe sectaries, thisaccuracy wassecured:
forms oftext, lists ofpeculiarities and treatises upon them,
and soon.When thiskind ofcarebeganinthecase of
eachtext, andwhat oforiginalcharacter mayhave been
effaced beforeit,orlostinspiteofit,cannot betold. But
itiscertain that theVedic recordsfurnish, onthewhole,
awonderfullyaccurate andtrustworthy pictureofaform of
ancient Indianlanguage (aswell asancient Indian beliefs
andinstitutions) which was anatural andundistorted one,
andwhichgoesback agoodwaybehind theclassical San-
skrit. Itsdifferences from thelatter thefollowingtreatise
endeavors toshow indetail.
Along with theverses and sacrificial formulas and
phrasesinthetexts oftheBlackYajur-Vedaaregiven
long prose sections,inwhich theceremonies aredescribed,
theirmeaning andthereason ofthedetails andtheaccom-
panying utterances arediscussed andexplained,illustrative
legendsarereportedorfabricated, andvariousspeculations,
etymological andother, areindulgedin.Such matter comes
tobecalled brahmana(apparently 'relatingtothebrahman
orworship';.IntheWhite Yajur-Veda.itisseparated into
|
Summarize page 12: | xvi' INTRODUCTION.
awork by itself, beside thesamhitci ortext ofverses and
formulas, and iscalled theCatapatha-Brahmana, 'Brahmana
ofahundredways'.Other similar collections arefound, be-
longingtovarious other schools ofVedicstudy, andthey
bear thecommon name ofBrahmana, with thename ofthe
school, orsome other distinctivetitle, prefixed. Thus, the
AitareyaandKamhitaki- Brahmanas, belongingtothe
schools oftheRig-Veda,thePancavinqa andShadvin$a-
Brahmanas andother minor works,totheSama-Veda; the
Gopatha-Brahmana,totheAtharva-Veda;andaJaimini-
Brahmana, totheSama-Veda, hasjust (Burnell) been dis-
covered inIndia; theTaittirlya-Brahmanaisacollection
ofmingledmantra andbrahmana, like thesamhita ofthe
same name, butsupplementaryand later. These works are
likewise regardedascanonical bytheschools, andarelearn-
edbytheir sectaries with thesame extreme carewhich is
devoted tothesamhitas, and their condition oftextual
preservationisofakindred excellence. Toacertain
extent, there isamong them thepossessionofcommon
material: afact thebearingsofwhich arenotyetfully
understood.
Notwithstandingtheinanityofnosmallpartoftheir
contents, theBrahmanas areofahighorder ofinterest in
theirbearings onthehistoryofIndian institutions; and
philologically theyarenot lessimportant,sincetheyre-
presentaform oflanguageinmostrespects intermediate
between theclassical andthat oftheVedas, and offerspe-
cimens onalargescale ofaprose style,andofonewhich
isinthemain anatural andfreely developed one the
oldest andmostprimitive Indo-European prose.
Beside theBrahmanas aresometimes found laterap-
pendices,ofasimilarcharacter, called Aranyakas ('forest-
sections'): astheAitareya-Aranyaka, Taittirtya-Aranyaka,
Brhad-Aranyaka, and soon.Andfromsome ofthese, or
even from theBrahmanas, areextracted theearliest Upa-
nishads('sittings,lectures onsacredsubjects') which,
howr
ever, arecontinued andadded todown toacompara-
tively modern time. TheUpanishadsareoneofthe lines
|
Summarize page 13: | INTRODUCTION. xvii
bywhich theBrahmana literaturepassesover into thelater
theologicalliterature.
Another line oftransition isshown intheSutras
(-lines,
rules'). Theworks thusnamed areanalogous with the
Brahmanas inthatthey belongtotheschools ofVedic
studyand arenamed from them, andthattheydealwith
thereligiousceremonies:treating them, however, inthe
wayofprescription,notofdogmatic explanation. They,
too, contain some mantra orhymn-material, notfound to
occur elsewhere. Inpart ($rautaorkalpa-sutras\ theytake
upthegreatsacrificial ceremonies, with which theBrah-
manas have todo;inpart (grhya-sutras), theyteach the
minor duties ofapious householder; insome cases(sa-
mayacarika-sutras) they laydown thegeneral obligationsof
onewhose life isinaccordance withprescribed duty. And
outofthelasttwo, orespeciallythe last,come bynatural
developmentthelaw-books(dharma-$astras), which make
aconspicuous figureinthelater literature :theoldest and
most noted ofthem beingthat called bythename of
Manu(anoutgrowth,itisbelieved, oftheManava Vedic
school);towhich areadded thatofYajnavalkya. andmany
others.
Respectingthechronologyofthisdevelopment,orthe
date ofanyclass ofwritings,stillmore ofanyindividual
work, the less that issaidthebetter. Alldatesgivenin
Indianliterary historyarepinssetuptobebowled down
again. Every important work hasundergonesomany more
orlesstransforming changesbeforereachingtheform in
which itcomes tous,that thequestionoforiginal con-
struction iscomplicatedwith that offinal redaction. Itis
sowith thelaw-book ofManu, justmentioned, which has
well-founded claims tobeing regardedasoneofthevery
oldest works oftheproperSanskrit literature,ifnotthe
oldest(itisvariously assigned,toperiodsfrom sixcenturies
before Christ tosoon afterChrist).Itisso,again,ina
stillmorestriking degree,with thegreat legendary epicof
theMahdbharata. Theground-workofthis isdoubtless of
very early date; but ithasserved asatext intowhich
b
|
Summarize page 14: | xviiiINTRODUCTION.
materials ofvarious character andperiod havebeeninwoven,
until ithasbecome aheterogeneous mass, akind ofcyclo-
pediaforthewarrior-caste, hard toseparateinto itscon-
stituent parts.ThestoryofNala, and thephilosophical
poem Bhagavad-Glta,aretwo ofthemost noted ofits
episodes.TheRamayana,theother most famousepic,
isawork ofanother kind: thoughalsoworked overand
more orless altered initstransmission toourtime,itis
theproduction,inthemain, ofasingleauthor(Valmiki);
and itisgenerallybelieved tobeinpart allegorical,re-
presentingtheintroduction ofAryanculture anddominion
intoSouthern India. Byitssidestand anumber ofminor
epics,ofvarious authorship andperiod,astheRaghuvah$a
(ascribedtothedramatistKalidasa),theMaghakavya,the
Bhattikavya (the last, writtenchieflywith thegrammatical
intent ofillustrating byuse asmanyaspossibleofthe
numerous formations which, through taught bythegram-
marians,findnoplaceintheliterature).
ThePurdnas. alargeclass ofworksmostlyofimmense
extent, arebestmentioned inconnection with theepics.
Theyarepseudo-historical andpropheticincharacter, of
modern date, andofverysmall value. Realhistoryfinds
noplaceinSanskritliterature, nor isthere anyconscious
historical element inanyoftheworkscomposingit.
Lyric poetryisrepresented bymany works, some of
which, astheMeghaduta andGitagovinda, areofnomean
order ofmerit.
Thedrama isastillmorenoteworthy andimportant
branch. The first indications ofdramatical inclination and
capacity onthepartoftheHindus areseen incertain
hymnsoftheVeda, where amythologicalorlegendary
situation isconceiveddramatically, and setforth inthe
form ofadialogue well-knownexamplesarethedialogue
ofSarama andthePanis, that ofYama and hissister Yami,
that ofVasishtha andtherivers, that ofAgni andtheother
gods butthere arenoextant intermediaries between these
andthestandard drama. Thebeginningsofthelatter date
from aperiod when inactual lifethehigher andeducated
|
Summarize page 15: | INTRODUCTION. xix
characters used Sanskrit, andthelower anduneducated used
thepopulardialects derived fromit,thePrakrits; and their
dialoguereflects this condition ofthings. Then, however,
learning (nottocall itpedantry)intervened, andstereotyped
thenew element; aPrakrit grammar grew upbeside the
Sanskrit grammar, accordingtotherules ofwhich Prakrit
could hemadeindefinitely onasubstrate ofSanskrit; and
none oftheexisting dramas need todatefrom thetime of
vernacular useofPrakrit, while most orallofthem are
undoubtedly much later. Amongthedramatic authors,
Kalidasa isincomparablythe chief, and hisCakuntala as
distinctlyhismasterpiece.His datehasbeen amatter of
muchinquiry andcontroversy;itisdoubtless some cen-
turies later than our era.Theonlyother work deserving
tobementionedalongwith Kalidasa' sistheMrchakafl of
Cudraka, also ofquestionable period,butbelieved tobe
theoldest oftheextant dramas.
Apartlydramatic character belongsalso tothefable,
inwhich animals arerepresentedasactingandspeaking.
Themost noted works inthisdepartmentarethePanca-
tantra, whichthroughPersian andSemitic versions hasmade
itswayallover theworld, and contributes aconsiderable
quotatothe fable-literature ofevery European language,
and, partly founded onit,thecomparativelyrecent and
popular Hitopade$a ('salutary instruction').
Two oftheleading departmentsofSanskrit scientific
literature, thelegal andthegrammatical,havebeen already
sufficiently noticed; ofthoseremaining,themost important
byfar isthephilosophical. Thebeginningsofphilosophic-
alspeculationareseenalreadyinsome ofthelaterhymns
oftheVeda, moreabundantlyintheBrahmanas andAran-
yakas, andthenespeciallyintheUpanishads.Theevo-
lution and historic' relation ofthesystemsofphilosophy,
andtheageoftheir text-books, arematters onwhichmuch
obscuritystill rests. There aresixsystemsofprimary rank,
andreckoned asorthodox, although really standinginno
accordance with approved religiousdoctrines. Allofthem
seek thesame end, theemancipationofthesoulfrom the
b*
|
Summarize page 16: | xx INTRODUCTION.
necessityofcontiiiuingitsexistence inasuccession of
bodies, and itsunification with theAll-soul; butthey
differ inregardtothemeans bywhichtheyseek toattain
thisend.
Theastronomical science oftheHindus isareflection
ofthat ofGreece, and itsliterature isofrecent date; but
asmathematicians, inarithmetic andgeometry ,theyhave
shown more independence.Their medical science, although
itsbeginnings goback even totheVeda, intheuseof
medicinalplantswithaccompanyingincantations, isoflittle
account, and itsproperliterature bynomeans ancient.
|
Summarize page 17: | CONTENTS.
Chap.
PREFACE
INTRODUCTION
I.ALPHABET
II.SYSTEM OFSOUNDS :PRONUNCIATION .
Vowels, 8;Consonants, 11;Quantity, 26;Accent, 27.
HI.RULES OFEUPHONIC COMBINATION . .
Introductory, 33;Principles, 36;Rules ofVowel Com-
bination, 41;Permitted Finals, 46;Deaspiration, 50;
Surd andSonant Assimilation,51
;Combinations of
Final sandr,53;Conversion ofstos,57
;Con-
version ofnton,60;Conversion ofDental Mutes to
Linguals and Palatals, 62;Combinations ofFinal n,
63;Combinations ofFinal m,65;thePalatal Mutes
and Sibilant, andh,66;theLingual Sibilant, 71;
Extension andAbbreviation, 72;Strengthening and
Weakening Processes, 73;Guna and Vrddhi, 74;
Vowel-lengthening, 76;Vowel-lightening, 77
;Nasal
Increment, 78; Reduplication, 79.
IV.DECLENSION
Gender, Number, Case, 80:Uses oftheCases, 81;
Endings ofDeclension, 92; Variation ofStem, 95;
Accent inDeclension, 97.
V.NOUNS ANDADJECTIVES
Classification etc.,99;Declension I.,Stems ina,100;
DeclensionII.,Stems in iandu,104; Declension
III., Stems inLong Vowels(a, I,u):A.Root-words
etc., Ill; Stems inDiphthongs, 116; B.Derivative
Stems etc.117; Declension IV., Stems inrorar,
123; Declension V.,Stems inConsonants, 127;
A.Root-stemsetc., 129; B.Derivative Stems inas,
is,us,138; C.Derivative Stems inan,140; D.,
inin,145; E.,inantorat,146; F.Perfect Par-
ticiples invans, 152; G.Comparativesinyas,155;
Comparison,156.Page.
V
ix
18
832
3379
8098
99159
|
Summarize page 18: | xxiiCONTENTS.
Chap.
VI.NUMERALS 160167
Cardinals, 160; Ordinals etc., 166.
VIE.PRONOUNS 168 181
Personal, 168
;Demonstrative, 171;Interrogative,
176; Relative, 177;Emphatic, 179;Nouns used pro-
nominally, 179; Pronominal Derivatives, Possessives
179; Adjectives declined pronominally, 181.
VIII. CONJUGATION 182 206
Voice, Tense, Mode, Number, Person, 182; Verbal
Adjectives andNouns, 185; Secondary Conjugation,
185; Personal Endings, 186;Subjunctive Mode, 191;
Optative, 193;Imperative, 195;Uses oftheModes,
196;Participles, 201;Augment, 201;Reduplication,
202; Accent oftheVerb, 203.
IX.THEPRESENT-SYSTEM . ;.[,,,.T.~^J-rt^>207 255
General, 207; Conjugations andConjugation Classes,
208;I.Root-class (second orad-class), 211;II.Re-
duplicating Class (thirdorftu-class), 221;III.Nasal
Class (seventhorrwd/i-class), 229; IV.Nuandw-Classes
(fifthandeighth,orsuandtan-classes), 232; V.Na-
Class (ninth orfcrz-class), 238; VI.a-Class(first or
6ftu-classj, 241;VII.Accented d-Class(sixth ortud-
class), 245; VIII. Fa-Class(fourth ordiu-elass), 248;
IX.Accented yd-Class orPassiveConjugation, 252;
Uses ofthePresent andImperfect, 254.
X.THEPERFECT-SYSTEM 255 270
Perfect Tense, 255; PerfectParticiple, 266; Modes
ofthePerfect, 267; Pluperfect, 269; Uses ofthe
Perfect, 270.
XI.THEAORIST-SYSTEMS 271 299
Classification, 271;I.Simple Aorist: 1.Root-aorist,
273
4;Passive Aorist 3dsing., 277; 2.thea-Aorist,
278; II. 3.Reduplicated Aorist, 281; III. Sibilant
Aorist, 285; 4.thes-aorist, 286; 5.the is-Aorist,
290; 6.the>-aorist, 293;7.thesa-Aorist, 294;
Precative, 296; Uses oftheAorist, 298.
XII.THEFUTURE-SYSTEMS 299 307
I.Thes-future, 300; Modes ofthe5-future, 302;
Participles ofthes-future, 302; Preterit ofthe s-
future:Conditional, 303; II.ThePeriphrastic Future,
303
;Uses oftheFutures andConditional, 305.
|
Summarize page 19: | CONTENTS. xxiii
Chap. Page.
XIII. VERBAL ADJECTIVES ANDNOUNS :PARTICI-
PLES, INFINITIVES, GERUNDS 307 321
PassiveParticipleintaornd,307;Past Active Par-
ticipleintavant, 310; Future PassiveParticiples:
Gerundives, 310; Infinitives, 313; Uses oftheInfini-
tives, 315; Gerunds, 319; Adverbial Gerund inam,
321.
XIV. DERIVATIVE ORSECONDARY CONJUGATION .321 347
I.Passive, 322;II.Intensive, 323; Present-System,
325; Perfect.Aorist, Future, etc.,329; III.Desider-
ative, 331; Present-System, 334; Perfect, Aorist,
Future, etc.,335; IV.Causative, 337; Present-System,
339; Perfect, Aorist, Future, etc.,340; V.Denomi-
native, 343.
XV.PERIPHRASTIC ANDCOMPOUND CONJUGATION 347 357
The Periphrastic Perfect, 347; Participial Periphras-
ticPhrases, 349;Composition withPrepositional
Prefixes, 350; Other Verbal Compounds, 355.
XVI. INDECLINABLES 357 370
Adverbs, 358;Prepositions, 366;Conjunctions, 369;
Interjections, 369.
XVH. DERIVATION OFDECLINABLE STEMS . . .370 424
A.Primary Derivatives, 373; B.Secondary Deriva-
tives, 403.
XVIII. FORMATION OFCOMPOUND STEMS.... 424 456
Classification, 425;I.Copulative Compounds, 428;
II.Determinative Compounds, 431;A.Dependent
Compounds, 432; B.Descriptive Compounds, 437;
III.Secondary Adjective Compounds, 443; A.Pos-
sessive Compounds, 443; B.Compounds withGoverned
Final Member, 452; Adjective CompoundsasNouns
andasAdverbs, 453; Anomalous Compounds, 455;
Stem-finals altered inComposition, 455;Irregular
Construction withCompounds, 456.
APPENDIX 457 460
A.Examples ofVarines Sanskrit Type457;B.Exam-
pleofAccentuated Text, 459.
SANSKRIT INDEX 461 475
GENERAL INDEX 476485
|
Summarize page 20: | XXIVABBREVIATIONS.
ABBREVIATIONS.
AB.Aitareya-Brahmana.
APr.Atharva-Prati^akhya.
AV.Atharva-Veda.
BB. Bohtlingk andRoth (Petersburg
Lexicon).
9or ak.Qakuntala.
$B.$atapatha-Brahmana.
QGS. Qankhayana-Grihya-Sutra.
GB.Gopatha-Brahmana.
H.Hitopade$a.
K.Kathaka.
KB.Kaushitaki-Brahmana.
KSS. Katha-Sarit-Sagara.
M.Mann.
MBh. Mahabharata.
Megh. Meghadiita.MS. Maitrayani-Sauhita.
PB.Pancavin^a-Brahmana.
R.Ramayana.
Ragh. Raghuvan^a.
RPr.Rigveda-Prati^akhya.
RV.Rig-Veda.
SB.Shadvin^a-Brahmaua.
SV.Sama-Veda.
TA. Taittiriya-Aranyaka.
TB.TaittirTya-Brahmana.
TPr.Taittiriya-Prati^akbya.
Tribh. Tribhashyaratna.
TS.TaittirTya-Sanhita.
V.Veda.
VPr.Vajasaneyi-Prati^akhyj
VS.Vajasaneyi-Sanhita.
|
Summarize page 21: | CHAPTER I.
ALPHABET.
1.THE natives ofIndia write their ancient andsacred
languageinavarietyofalphabets generally,ineach
partofthecountry,inthesamealphabet whichtheyuse
fortheirown vernacular. Themode ofwriting, however,
which isemployed throughouttheheart ofAryan India, or
inHindustanproper,isalone adopted byEuropeanscholars :
itiscalled thedevanagari.
Thisname isofdoubtful origin and value. Amorecomprehensive name
isnagarl (perhaps,'ofthecity'); anddeva-nagarl is'nagarl ofthegods,'
or'oftheBrahmans.'
2.Much that relates tothehistory oftheIndianalphabetsisstill
obscure. The earliest written monuments ofknown date inthecountry are
theinscriptions containing the edicts ofA^okaorPiyadasi, ofabout the
middle ofthethird century B.C.They areintwo different systems of
characters,ofwhich oneshows distinct signsofderivation from aSemitic
source, while theother isalsoprobably, though much less evidently, ofthe
same origin (Burnell). From thelatter, theLath, orSouthern Acoka cha-
racter(ofGirnar), come thelater Indian alphabets, both those ofthenorthern
Aryan languages, and those ofthesouthern Dravidian languages. The
nagari, devanagari, Bengali, Guzerati, andothers,arevarieties ofitsnorthern
derivatives; andwith them arerelated some ofthealphabets ofpeoples
outside ofIndia asinTibet andFarther India whohaveadopted Hindu
culture orreligion.
There isreason tobelieve that writing was firstemployedinIndia for
practical purposesforcorrespondence andbusiness and thelike and
onlybydegrees came tobeappliedalso toliterary use. Theliterature,to
agreat extent, andthemore fully inproportiontoitsclaimed sanctity and
authority, ignoresallwritten record, andassumes tobekeptinexistence by
oral tradition alone.
3.Ofthedevanagari itself there areminor varieties, depending on
differences oflocality orofperiod,asalso ofindividual hand(seeexamples
Whitney, Grammar. 1
|
Summarize page 22: | I.ALPHABET.
inWeber's catalogueoftheBerlin Sanskrit MSS., inRajendralala Mitra's
notices ofMSS. inIndian libraries, inthepublished fac-similes ofin-
scriptions, and soon);andthese areinsome measure reflected inthetype
preparedforprinting,both inIndia and inEurope. But astudent who
makes himself familiar with one style ofprintedcharacters willhave little
difficulty with theothers, and willsoonlearn, bypractice,toreadthemanu-
scripts. Afewspecimensoftypesother than those used inthiswork will
begiveninanAppendix.
Onaccount ofthe difficulty ofcombining them with thesmaller sizes
ofourRoman and Italictype,thedevanagari characters willbeused below
only inconnection with the first orlargest size. And, inaccordance with
thelaudable usageofrecent grammars, they will, wherever given, bealso
transliterated initalic letters; while the latter alone willbeused inthe
other sizes.
4.Thestudent maybeadvised totrytofamiliarize himself
from the start with thedevanagarl mode ofwriting. Atthesame
time,itisnotnecessarythatheshould dosountil, having
learned theprincipal paradigms, hecomes tobegin reading and
analysing andparsing;andmanywill find thelatter themore
practical, andintheendequallyormoreeffective, way.
5.The characters ofthedevanagarl alphabet, andthe
Europeanletters which willbeused intransliterating them,
areasfollows :
short.long.
Vowels :
simplepalatal
labial
lingual
dental
diph-(palatal
thongsIlabial
Visarga
Anusvara$l
3u
^rd*
FTI
^e
37ou
f
1}
ai
au
,norm(see 73)
Mutessurd
guttural
palatal
lingual
dental
labial qpsurd asp.
is1^kh
233TCh
2S%th
335Tth
ssCfiphsonant
19*Tg
24Sfj
293"d
34$d
39STbson. asp.
oqgh
53Tjhnasal
263Tn
dh 31HIn
dh se^n
bh 4iITm
|
Summarize page 23: | palatal
|
Summarize page 24: | 4I.ALPHABET.
alphabeticscheme above areusedonlywhen thevowel
forms asyllable byitself, orisnotcombined with apre-
cedingconsonant: thatis,when itisinitial, orpreceded
byanother vowel. Incombination with aconsonant, other
modes ofrepresentationareused.
B.Ifmore consonants than oneprecedethevowel,
formingwith itasingle syllable,their characters must be
combined into asingle compoundcharacter.
Ordinary Hindu usagedoes notdivide thewords ofasen-
tence, anymore than thesyllablesofaword;afinalconsonant
iscombined intoonesyllablewith the initial vowel orconso-
nant ofthenext following word.
10.Under A, itistobenoticed that themodes of
indicatingavowel combined with aprecedingconsonant
areasfollows:
a.The short^ahasnowrittensignatall:thecon-
sonant-signitselfimpliesafollowingTa,unless some other
vowel-signisattached toit(orelsethevirama:11).Thus
theconsonant-signsasgivenabove inthealphabetic scheme
arereallythesignsofthesyllables ka,kha, etc. etc.(toha].
b.The-long ^Taiswrittenbyaperpendicularstroke
after theconsonant: thus, 5fiTka, EfTdha,^Tha.
c.Short^iandlong ^e,byasimilarstroke, which
forshort iisplaced before theconsonant and forlongIis
placedafterit,and ineither case isconnected with the
consonant byahook above theupperline:thus, f^R ki,
3ft ki',Prbhi,Htbhi
;ftm', jftnl.
Thehookabove, turning tothe left ortotheright,ishistorically the
essentialpart ofthecharacter, having beenoriginally thewhole ofit;the
hooks were only laterprolonged, soastoreach allthewaydown beside
theconsonant. IntheMSS., they almost never have thehorizontal stroke
drawn across themabove, though this isadded inalltheprintedforms of
thecharacters*.
*Thus, originally dfiW,ofjit;intheMSS., jcfj, effj ;inprint,
|
Summarize page 25: | 12] WRITING OFVOWELS. 5
d.Thew-sounds, short andlong,arewritten byhooks
attached tothelower end oftheconsonant-sign: thus,Sfj
ku, 3\ku;Idu,Idu.Onaccount ofthenecessities of
6\ O SX
combination, duandduaresomewhatdisguised: thus, If.
^;andtheforms with^rand^Thare stillmore irre-
gular:thus,"^\ru,%ru;<^Thu,f^1hu.
e.The r-vowels, short andlong,arewritten byasub-
joined hook, singleordouble,opening toward theright:
thus, ^\kr, Sfjkf ;dr,^dr.Inthe/j-sign,thehooks
areusuallyattached tothemiddle: thus,^hr,^hr.
Astothecombination ofrwith preceding r,seebelow, 14.
f.The/-vowel iswritten with areduced form ofits
full initial character: thus,efikl:thecorresponding long
hasnorealoccurrence(23),butwould bewritten with a
similar reducedsign.
g.Thediphthongsarewritten bystrokes, singleor
double, above theupper line, combined, for Jtoand ETF
aUjwith the#-signafter theconsonant :thus. 3ftke,fi
kai;^tko,
Insome devanagari MSS.(asintheBengali alphabet),thesingle stroke
above,orone ofthedouble ones,isreplaced byasign like thea-sign
before theconsonant :thus, (off fee, \fi\ feat, fofil feo, |cftl feaw.
11.Aconsonant-sign, however,iscapableofbeing
made tosignifytheconsonant-sound alone, without an
added vowel, byhavingwritten beneath itastroke called
thevirama('rest, stop'): thus,fik^<d,^h.
Since, aswas pointed outabove, theHindus write thewords ofa
sentence continuously,likeoneword(9,end),thevirama isingeneral called
foronlywhen afinalconsonant occurs before apause. But itisalsooccasion-
allyresorted tobyscribes,orinprint,inorder toavoid anawkward or
difficult combination ofconsonant-signs; and itisused freely inpublished
texts which fortheconvenience ofbeginners have their words printed sepa-
rately.
12.Under B,itistobenoticed that theconsonant
combinations areforthemostpartnot atalldifficult to
|
Summarize page 26: | 6I.ALPHABET. [12
make ortorecogniseforonewho isfamiliar with the
simple signs.The characteristicpartofaconsonant-sign
that istobeadded toanother istaken(totheexclusion
ofthehorizontal oroftheperpendicular framing-line,or
ofboth), andtheyareputtogether accordingtoconveni-
ence,either sidebyside, oroneabove theother: insome
combinations either arrangementisallowed. Theconsonant
that istobepronouncedfirst issetbefore theother inthe
one order, andabove itintheother order.
Examplesoftheside-by-side arrangementare :ITfgga,
STjja, C?jpya, ZTJnma, f^Ttiha,^Tbhya, F^Iska, STUma,
f3Ttka.
Examplesoftheabove-and-belowarrangement are :
^TMa, sfcca,^nja,^dda,Hpta,^tna.
13.Insomecases, however, there ismore orless
abbreviation ordisguiseoftheindependent form ofacon-
sonant-signincombination.
Thus, of3\kin"^7kid,^\Ida;and in^HTknaetc.;
ofrTtinfFtta;
of^din"%dga,^dna, etc.;
of*Tmand ZTy,whenfollowing other consonants :
thus,^Tkya,3Rkrna,^nma, 2Tnya,^dma, t%dya,^T
hma, ^Thya."5Tchya,^dhya ;
of5Tp,whichgenerally becomes $Twhen followed
byaconsonant:thus, 31pea,Wpraa, ^j$va, V3J$ya.The
samechangeisusualwhen avowel-signisadded below:
thus, 5Tpw, 5Tcr.O (.
Othercombinations, ofnotquite obviousvalue, are
STnna, ^T//a,^ddha,^dbha,^sta, Tgstha; andthe
compounds of^h:as^r7m,^A^a.
Inacase ortwo,notrace oftheconstituent letters is
recognisable:thus,^ksa,^jfia.
14.Thesemivowel^r,inmaking combinations with
|
Summarize page 27: | 16]COMBINATIONS OFCONSONANTS. 7
other consonants,istreated inawholly peculiar manner,
analogouswith that ofthevowels. Ifpronouncedbefore
another consonant (orconsonant-combination),itiswritten
with ahook above, openingtotheright (likethesubjoined
signofr:10e):thus. Rrka,^rsa(fP rtsna).Ifpro-
nounced after another consonant (aloneorincombination),
itiswritten with aslantingstroke below: thus, Elgra,
Pfpra,Tsra(and CETgrya,Tsrva); and, with modifica-
tions ofthepreceding consonant-signlikethose noted above,
"5Ttra, "5Tpr, ?Tdra.
When}Tristobecombined with afollowing5fJr,it
isthevowel which iswritten infull,with itsinitial char-
acter, and theconsonant insubordination toit:thus,
ftrr.
15.Further combinations, ofthree,orfour,oreven
fiveconsonant-signs,aremade accordingtothesame rules.
Examplesare:
ofthree consonants, f[ttva, 1STddhya, $3dvya,^
drya,SETdhrya,t3psva, 5JT9^ya, ^Jstya,^1hvya;
offour consonants, ^Jktrya, ^Tnksya, ^Jstrya,
rF?I tsmya;
offiveconsonants, fF^U rtsnya.
The manuscripts, andthetype-fontsaswell,differ from oneanother
more intheir managementofconsonant combinations than inany other
respect,often having peculiarities which oneneeds alittle practicetounder-
stand. Itisquiteuseless togiveinagrammar thewhole series ofpossible
combinations (manyofthem excessively rare) which areprovidedforinany
given type-font,oreven inall.There isnothing which due familiarity
with thesimple signs andwith theabove rules ofcombination willnotenable
thestudent toanalyse andexplain.
16.Asigncalled theavagraha ('separator') namely,
vl--isused inthemanuscripts,sometimes inthemanner
ofahyphen, sometimes asamark ofhiatus, sometimes to
mark theelision ofinitial 51aafter final^eorsqjo(135).
Inprintedtexts,especially European,itisordinarilylimited
|
Summarize page 28: | I.ALPHABET. [16
totheuselastmentioned: thus. ?t^^f^telbruvan, HT
so'bravit, for teabruvan, soabramt.
The signisused tomark anomission ofsomething.
Insome texts, ithasalso thevalue ofahyphen.
SignsofpunctuationareIand II.
17.Thenumeralfiguresare
11,^2,\3,94,H5,|6,b7.TT8.$9,0.
Incombination, toexpress largernumbers,theyare
used inpreciselythesamewaywithEuropean digits:thus,
^H25,^0630, <(000 1000, ^T7b 1879.
18.TheHindu grammarianscallthedifferent sounds, and
thecharacters representing them, byakara('maker')added to
thesound ofthe letter,ifavowel, ortotheletter followed by
a,if-aconsonant. Thus, thesound orcharacter aiscalled
akara; kiskakara: and soon.Butthekara isalso omitted,
and a,ka,etc. areused alone. Ther,however, isnever called
rakara, butonlyraorrepha ('snarl':theonly exampleofa
specific name foranalphabetic element ofitsclass). The ami-
svara andmsargaarealsoknown bythese names alone.
CHAPTER II.
SYSTEM OFSOUNDS; PRONUNCIATION.
I.Vowels.
19.THE a,i,and^^-vowels. TheSanskrit hasthese
three earliest andmost universal vowels ofIndo-European
language,inboth short andlongform %aand TTa,
^tand^I.3uand3Tu.Theyaretobepronouncedin
the"Continental" or"Italian" manner asinfarorfather,
pinandpique, pulland rule.
20.The aistheopenest vowel, anutterance from theex-
|
Summarize page 29: | 24] VOWELS. 9
pandedthroat;itstands, therefore., innorelation ofkindred
with anyoftheclasses ofconsonantal sounds. The ianduare
close vowels, made withmarkedapproachofthearticulating
organstooneanother :iispalatal, andshades through yinto
thepalatal andguttural consonant-classes;uissimilarly related,
through v,tothelabialclass, asinvolvinginitsutterance a
narrowing androundingofthelips.
ThePaninean scheme (commentarytoPanini's grammar,i.1.9)classes a
asguttural, butapparently only inorder togive that series aswell asthe
rest avowel: noone ofthePraticakhyas putsaintoone class with ketc.
Allthese authorities concur incalling the iandw-vowelsrespectively palatal
and labial.
21.The short aisnotpronouncedinIndia with the full
opennessofa,asitscorresponding short, butusuallyasthe
"neutral vowel"(Englishso-called "short",ofbut, son, blood,
etc.). Thispeculiarity appears very early, being acknowledged
byPanini andbytwo ofthePraticakhyas (APr.i.36;VPr. i.
72),which call theutterance samvrta, 'coveredup,dimmed'.
Itis,however, ofcourse notoriginal ;and itisjustly wont to
beignored byWestern scholars(exceptthosewhohave studied
inIndia).
22.Thea-vowels aretheprevailing vowel-sounds ofthe
language, being about twice asfrequentasalltheothers(in-
cluding diphthongs)takentogether. The -vowels, again,are
about twice asnumerous asthew-vowels. And, ineachpair,
theshort vowel ismore than twice(2y2to3times)ascommon
asthelong.
Formore precise estimates offrequency,ofthese and oftheother
alphabetic elements, and fortheway inwhich they were obtained, see
below, 75.
23.The rand/-vowels. Tothethree simplevo-
welsalready mentioned theSanskrit addstwo others, the
r-vowels and the/-vowel, both ofthemplainly generated
bytheabbreviation ofsyllables containinga^"ror^T/
alongwith another vowel :the TRrcoming (almost always:
see237, 241-3) fromT|"arorfra,theFT/from FTal.
Some oftheHindu grammarians add tothealphabetalso alongI;
but this isonly forthesake ofanartificial symmetry, since thesound does
notoccur inasingle genuine word inthelanguage.
24.Thevowel :fjrissimplyasmooth oruntrilled
r-sound, assumingavocalic office insyllable-making
|
Summarize page 30: | 10II.SYSTEM OFSOUNDS. [24
as,byalike abbreviation,ithasdone also incertain Sla-
vonic languages.Thevowel FT Iisan-sound similarly
uttered liketheEnglish/-vowel insuchwords asable,
angle,addle.
Themodern Hindus pronouncethese vowels as ri, ri,li
;orevenIri), having longlost thehabit and thefacilityof
givingavowel value tothepurerand^-sounds. Their example
iswidelyfollowed byEuropeanscholars;andhence also the
(distortingandquite objectionable)transliterationsri,n,li.
There isno-real difficultyinacquiringandpractisingthetrue
utterance.
Some ofthegrammarians (seeAPr. i.37,note) attempttodefine more
nearly theway inwhich,inthese vowels, areal ror^-element iscombined
with somethingelse.
25.Like their corresponding semivowels, randI,these
vowels belong respectivelyinthegeneral lingual anddental class-
es
;theeuphonicinfluence ofrandf(180)shows thisclearly.
Theyaresoranked inthePaninean scheme;butthePraticakhyas
ingeneral strangelyclassthem with thejihvcimuliya sounds, our
"gutturals".
26.The short risfound inevery varietyofword andof
position, and isnotrare, being justabout asfrequentaslongu.
Long fisverymuch more unusual, occurring onlyincertain
pluralcases ofnoun-stems inr(374, 378).The /ismetwith
onlyinsome oftheforms and derivatives ofasinglenotvery
common verbal root(kip).
27.Thediphthongs. Ofthefourdiphthongs, two,
the^eand ETTo,areingreat part original Indo-European
sounds. IntheSanskrit, they wear theaspectofbeing
productsoftheincrement orstrengtheningof^iand3u
respectively; andtheyarecalled thecorresponding guna-
vowels tothelatter(seebelow, 235). The other two,^ai
and tau,arebytheprevalent andpreferable opinionheld
tobeofpeculiar Sanskritgrowth (thereisnocertain trace
ofthem tobefound even intheZend); theyare also in
generalresults ofanother andhigherincrement of^iand
3u,towhichtheyarecalled thecorrespondingvrddhi-
vowels(below, 235). But allarelikewise sometimesgene-
|
Summarize page 31: | 32 DIPHTHONGS. 1j
rated byeuphoniccombination(127); and TTo,especially,
iscommon asresult ofthealteration ofafinal *3Ras175).
28.The^eand 3TFoare,both inIndia and inEu-
rope, usually pronouncedastheyaretransliterated that
is,aslonge(English "long ",oreinthey]ando-sounds,
withoutdiphthongalcharacter.
Suchthey apparently already were totheauthors ofthe
Praticakhyas, which, while ranking them asdiphthongs\san-
dfa/afaara), giverulesrespectingtheirpronunciationinamanner
implying them tobevirtually unitary sounds. But their euphonic
treatment(131-4) clearly shows them tohave been still atthe
period when theeuphonic laws established themselves, asthey
ofcourse were attheirorigin,realdiphthongs,ai(a-f- *)and
au\a-\- tf).From them, onthesame evidence, theheavier or
vrddhi diphthongs weredistinguished bythelengthoftheir a-
element, asai(a-{- i]andau(a-\-u).
The recognisable distinctness ofthetwoelements inthevrddhi-diph-
thongs isnoticed bythePraticakhyas (seeAPr. i.40,note); buttherelation
ofthose elements iseither defined asequal,ortheaismade oflessquan-
titythan the iand u.
29.Thelighterorywwa-diphthongsaremuch more frequent
(6or7times)than theheavier orvrddhi- diphthongs, andthe
eand aithan the oandau(ahalfmore).Bothpairsare
somewhat more than half ascommon asthesimpleiandu-
vowels .
30.Thegeneral name given bytheHindu grammarianstothevowels
issuara, 'tone'; thesimple vowels arecalled samanaksara, 'homogeneous
syllable', andthediphthongsarecalled sandhyaksara,icombination-syllable'.
Thepositionoftheorgans intheir utterance isdefined tobeoneofopenness,
orofnon-closure.
Astoquantity and accent, seebelow. 76 ff.;80 ff.
II.Consonants.
31.TheHindu name for'consonant' isvyan/ana,'mani-
fester'. The consonants aredivided bythegrammariansinto
sparca,icontact' or'mute', antahstha, 'intermediate' orisemivowel',
andusman, 'spirant'. Theywillhere betaken upanddescribed
inthis order.
32.Mutes. Themutes, sparca,aresocalled asinvolving
acomplete closure orcontact\sparca], andnotanapproximation
|
Summarize page 32: | 12II.SYSTEM OFSOUNDS. [32
only,ofthemouth-organs bywhich theyareproduced. Theyare
divided into fiveclasses orseries(varga], accordingtotheorgans
andpartsoforgans bywhich thecontact ismade;andeach
series iscomposedoffivemembers, differing accordingtothe
accompanimentsofthecontact.
33.The fivemute-series arecalledrespectively guttural,
palatal, lingual (orcerebral), dental, and labial; andthey
arearrangedintheorder asjustmentioned, beginningwith
thecontact made furthest back inthemouth, comingfor-
ward frompointtopoint,andendingwith thefrontmost
contact.
34.Ineach series there aretwosurdmembers, two
sonant, andonenasal (whichisalsosonant):forexample,
inthelabial series,tj^pandVf^ph, 3^band^bh,and ?Tm.
Themembers arebytheHindu grammarians calledrespectively 'first',
'second', 'third', 'fourth1
,and 'last' or'fifth'.
Thesurd consonants areknown asaghosa, 'toneless', andthesonants
asghosavant, 'having tone'
;andthedescriptionsofthegrammarians arein
accordance with these terms. All alike recognise adifference oftone, and
notinanymanner adifference offorce, whether ofcontact orofexpulsion,
asseparating thetwogreat classes inquestion. That thedifference depends
onvivara, 'opening', orsarhvara, 'closure'(oftheglottis),isalsorecognised
bythem.
35.The firstandthirdmembers ofeach series arethe
ordinary correspondingsurdandsonant mutes ofEuropean
languages: thus,^kand\g, t^tand d,qpand^b.
36.Nor isthecharacter ofthenasalanymore doubtful.
Whatq^misto^pand3[^,or^ntot^tand <d,that
isalsoeach other nasal toitsown series ofmutes :asonant
expulsion intoandthrough thenose, while themouth-
organsareinthemute-contact.
TheHindu grammarians give distinctly this definition. The nasal
(anunasika, 'passing through thenose') sounds aredeclared tobeformed by
mouth andnosetogether; ortheirnasality (anunasikya)tobegiven them
byunclosure ofthenose.
37.Thesecond andfourth ofeach series areaspirates:
thus, beside thesurdmute^kwehave thecorresponding
|
Summarize page 33: | 38 ASPIRATE MUTES. 1$
surdaspirate ^kh,andbeside thesonantTf.^,thecorres-
pondingsonantaspirate %gh.Ofthese, theprecisechar-
acter ismore obscure and difficult.
That theaspirates,allofthem, arerealmutes orcontact sounds, and
notfricatives(likeEuropeanthandphandch, etc.),isbeyond question.
Itisalsonotdoubtful inwhatway thesurd M,forexample,differs
from theunaspiratedt:such aspiratesarefound inmany Asiatic languages,
andeven insome European:they involve theslipping-outofanaudible bit
offlatusoraspiration between thebreach ofmute-closure andthefollowing
sound, whatever itmay be.They areaccurately enough represented bythe
thetc., withwhich,inimitation oftheLatin treatment ofthesimilar ancient
Greekaspirates, weareaccustomed towrite them.
The sonant aspirates aregenerally understood and described asmade
inasimilar way, with aperceptibleft-sound after thebreach ofsonant
mute-closure. But there areinsuperabletheoretical difficulties intheway
ofacceptingthisexplanation ;andsome ofthebest phonetic observers(as
A.J.Ellis) deny thatthemodern Hindu pronunciationisofsuch acharacter,
anddefine theelement following themute asa"glottal buzz", rather,oran
emphasizedutterance ofthebeginning ofthesuceeding sound. Thequestion
isone ofgreat difficulty, anduponittheopinionsofthehighest authorities
are stillmuch atvariance. Sonant aspiratesare still inuse inIndia,in
thepronunciationofthevernacular aswell asofthelearned languages.
Bythe Pratic.akhyas, theaspiratesofboth classes arecalled sosman:
which might mean either 'accompanied byarush ofbreath' (taking usman
initsmore etymological sense), or'accompanied byaspirant' (below, 59).
Andsome authorities define thesurd aspiratesasmade bythecombination
ofeach surd non-aspirate with itsown corresponding surd spirant; andthe
sonantaspirates,ofeach sonantnon-aspirate with thesonant spirant, the
ft-sound(below, 65). But thiswould make thetwo classes ofaspirates of
quite diverse character, andwould alsomake ththesame asts,thasts,ch
ascf which isinanymeasureplausible only ofthe last. Panini hasno
name foraspirates ;thescheme given inhiscomment(toi.1.9)attributes
tothem mahaprana, 'great expiration1
,and tothenon-aspirates alpaprana^
'smallexpiration'.
Itisusualamong Europeanscholars topronounceboth
classes ofaspiratesasthecorresponding non-aspirateswith
afollowingh:forexample,2TthnearlyasinEnglishboat-
hook,mphasinhaphazard,Tdhasinmadhouse, and so
on.This is(aswehave seen above) confessedlyaccurate
onlyasregardsthesurdaspirates.
38.Thesonantaspirates are(intheopinionofmost),or
atleastrepresent, original Indo-European sounds, while thesurd
|
Summarize page 34: | 14II.SYSTEM OFSOUNDS. i38
aspiratesaregenerally regardedasaspecialIndian development.
Theformer aremore than twice ascommon asthelatter. The
unaspirated (non- nasal)mutes areverymuch more frequent
(5times)than theaspirates (forthespecial frequencyofbhand
original gh,see50and66);andamong them thesurds aremore
numerous (2Y2times)than thesonants. The nasals(chieflyn
andm)arenearlyasfrequentasthesurd non-aspirates.
Wetakeupnow theseveral mute-series.
39.Guttural series:^k,T^kh, JT^,\gh,3"n.
These aretheordinary Europeankand^-sounds,with their
corresponding aspiratesandnasal(the last, likeEnglish ng
insinging}.
The gutturalsaredefined bythePratic.akhyas asmade bycontact of
thebase ofthetongue with thebase ofthejaw, andthey arecalled, from
theformer organ, jihvamullya, 'tongue-root sounds'. ThePaninean scheme
describes them simplyasmade inthethroat(kantha). From theeuphonic
influence ofakonafollowings(below, 180), wemay perhapsinfer that
intheir utterance thetongue waswelldrawn back into thehinder mouth.
40.The kisbyfarthecommonest oftheguttural series,
occurring considerably more often than alltheother four taken
together.The nasal, exceptasstandingbefore oneoftheothers
ofthe series,isfound onlyasfinal(afterthe loss ofafol-
lowing k),and inaverysmall number ofwords.
41.TheSanskritgutturalseriesrepresents onlyaminority
ofIndo-European gutturals;these lasthave suffered more and
more general corruption thananyother class ofconsonants. By
processesofalteration which areprovedtohavebeguninthe
Indo-European period,since thesame words exhibit connected
changesalso inotherlanguages ofthefamily, thepalatal mutes,
thepalatalsibilantc,and theaspiration h,have come from
gutturals. Seethese various sounds below.
42.Palatal series :r(c,^ch, s?Fj,37jh,3Tn.This
whole series isderivative, being generated bythecorruptionof
original gutturals. The ccomes from anoriginalk asdoes
also, byanotherdegreeofalteration, thepalatalsibilant c(see
below, 64).The/,inlikemanner, comes from ag;butthe
Sanskrit jincludes initself twodegrees ofalteration, onecor-
respondingtothealteration ofktoc,theother tothat ofktog
(seebelow, 219 :intheZend, these twodegreesareheld dis-
tinctly apart). The cissomewhat morecommon than thej
(aboutasfour tothree). Theaspirate chisverymuch less fre-
quent (atenth ofc),andcomes from theoriginal groupsk.
Thesonantaspirate jhisexcessively rare(occurringbutonce
|
Summarize page 35: | 45] PALATAL ANDLINGUAL MUTES. 15
intheVedic texts, andnothalf-a-dozen times intheBrahma-
nas);where found, itiseither onomatopoeticorofanomalous
ornotIndo-European originiintheso-called rootujh,itcomes
fromjandh}.The nasal, n,never occurs except immediately
before or,inasmall number ofwords, also after(201)
oneoftheothers ofthe series.
43.Hence, intheeuphonic processesofthelanguage,the
treatment ofthepalatalsisinmany respects peculiar. Insome
situations, theoriginal unaltered guttural shows itself or,as
itappearsfrom thepointofview oftheSanskrit, thepalatal
reverts toitsoriginal guttural. Nopalatalever occurs asafinal.
Thejisdifferently treated, accordingasitrepresentstheone
ortheother degreeofalteration. And candj(except artificially,
inthealgebraicrules ofthegrammarians)donotinterchange,
ascorrespondingsurdandsonant.
44.Thepalatalmutes arebyEuropeanscholars, asby
themodern Hindus also, pronouncedwith thecompound
sounds ofEnglishchandj(inchurch andjudge}.
Their description bythe oldHindu grammarians, however, gives them
anot less absolutely simplecharacter than belongs tothe other mutes.
They arecalled talavya, 'palatal1
,and declared tobeformed against the
palate bythemiddle ofthetongue. They seem tohave been, then,
brought forward inthemouth from theguttural point, andmade against the
hard palateatapoint not farfrom thelingual one (below, 45), butwith
theupperflatsurface ofthetongue instead ofitspoint. Such sounds, in
alllanguages, pass easily into the(English)chand ./-sounds. The value
ofthechasmaking theprecedingvowel "long byposition" (227), and its
frequent origination from t-f-p(203),lead tothesuspicion that it.atleast,
mayhave had this character from thebeginning:compare 37,above.
45.Lingual series:t,3"th,Zd,?odh,HTn.The
lingual mutes arebyallthenative authorities denned as
uttered with thetipofthetongueturned upanddrawn
back intothedome ofthepalate (somewhat astheusual
English smooth rispronounced). Theyarecalled bythe
grammarians murdhanya, literally'head-sounds,capitals,
cephalics'; which term isinmany European grammars
rendered by'cerebrals' .Inpractice, among EuropeanSans-
kritists, noattemptismade todistinguish them from the
dentals : tispronouncedlike rTt,Jdlike
e["d,and so
w.ith the rest.
|
Summarize page 36: | 16II.SYSTEM OFSOUNDS. [46
46.The lingualsareanother non-originalseries ofsounds,
coming mainly from thephoneticalteration ofthenextseries,
thedentals, but also inpart occurringinwords thathave no
traceable Indo-European connection, andareperhapsderived from
theaboriginal languagesofIndia. Thetendencytolingualization
isapositiveoneinthehistoryofthelanguage:dentalseasily
passinto linguals under theinfluence ofcontiguousorneighbor-
inglingual sounds, butnotthecontrary;and allthesounds
oftheclassbecome markedly more frequentinthelater litera-
ture. The conditions oftheir ordinaryoccurrence arebriefly
these :a.scomes froms,much more rarely fromq,j,ks,in
euphoniccircumstances stated below(180, 218, etc.);b.a
dental mute followingsisassimilated toit,becoming lingual
(t,th, n);c.nisoften changedtonafter alingual vowel or
semivowel orsibilant inthesameword(189 etc.) ;d.dh,which
isofveryrare occurrence, comes from assimilation ofadental
after s(198 a)orh(222);e. tanddcomeoccasionally by
substitution forsome other sound which isnotallowed tostand
asfinal(142, 145).Whenoriginatedinthese ways,thelingual
letters mayberegardedasnormal;inanyother cases oftheir
occurrence, theyareeither productsofabnormalcorruption,or
signsofthenon-Indo-Europeancharacter ofthewords inwhich
they appear.
Inacertain number ofpassages numerically examined (below, 75),
theabnormal occurences oflingual mutes were less than half ofthewhole
number (74out of159), andmost ofthem(43)were ofn: allwere found
more frequentinthelater passages. IntheRig-Veda, only 15words have
anabnormal t;only 6,such ath;only 1,such adh;about 20(including
9roots, nearly allofwhich havederivatives) show anabnormald,besides
9thathavend;and30(including1root)show an.
Taken alltogether, thelingualsarebyfartherarest class
ofmutes(about iy2percent, ofthealphabet) hardlyhalf
asfrequent even asthepalatals.
47.Dental series:<^t,Z^tk, <[d,V^dk,^n.These
arecalledbytheHindus alsodantya, 'dental', and are
described asformed attheteeth(orattheroots ofthe
teeth), bythetipofthetongue. Theyarepracticallythe
equivalents ofourEuropean t,d,n.
Butthemodern Hindus aresaid topronouncetheir dentals with the
tipofthetongue thrust wellforward against theupper teeth, sothatthese
sounds getaslight tinge ofthequality belongingtotheEnglish andModern
Greek tft-sounds. Theabsence ofthatquality intheEuropean (especially
|
Summarize page 37: | 52]LABIAL MUTES; SEMIVOWELS 17
theEnglish) dentals isdoubtless thereason why tothe earofaHindu the
latterappearmore analogous with hislinguals, andhe isapttousethe
lingualsinwriting European words.
48.Thedentals areoneofthethreeIndo-European original
mute-classes. Intheir occurrence inSanskrittheyarejustabout
asfrequentasalltheother four classes takentogether.
49.Labial series: qpCfiph^b.^bh, 3}m.X^-X**X -X^^X
These sounds arecalledosthya,'labial1
,bytheHindu gram-
marians also. They are, ofcourse, theequivalents ofour
p,b,m.
50.Thenumerical relations ofthe labials arealittlepe-
culiar. Owingtotheabsence(oralmost entireabsence)ofbin
Indo-European,theSanskrit balso isgreatly exceeded infre-
quency bybh,which isthemostcommon ofallthesonant
aspirates,asphistheleastcommon ofthesurd. Thenasalm(not-
withstandingitsfrequent euphonic mutations when final: 212ff.)
occurs justabout asoften asalltheother fourmembers ofthe
series together.
51.Semivowels: fy,^"r,s$ I,3v.Thename given tothis
class ofsounds bytheHindu grammariansisantahstha, "standing between'
either from their character asutterances intermediate between vowel and
consonant, or(more probably) from thecircumstance oftheir being placed
between themutes and spirants inthearrangement oftheconsonants.
Thesemivowels areclearlyakinwith theseveral mute series
intheirphysical character, andtheyare classifiedalong with
those series though notwithout some discordances ofview
-bytheHindu grammarians. Theyaresaid tobeproduced
with theorgans "slightlyincontact"(isatsprsta),or"inimperfect
contact"(duhsprsta].
52.The^risclearly shown byitsinfluence inthe
euphonic processesofthelanguagetobealingual sound,
oronemade with thethetipofthetongue turned upinto
thedome ofthepalate.Itthus resembles theEnglish
smoothr,and,likethis,, seems tohavebeen untrilled.
ThePaninean scheme reckons rasalingual. None ofthePraticakhyas,
however, does so;norarethey entirely consistent with oneanother inits
description. Forthemostpart, they define itasmade at'the roots ofthe
teeth'. Thiswould giveitapositionlikethat ofthevibrated r;butnoau-
thority hints atavibration asbelongingtoit.
Whitney, Grammar. 2
|
Summarize page 38: | 1II.SYSTEM OFSOUNDS.[52
Inpointoffrequency,rstands very highonthe listof
consonants ;itisabout equalwith v,n,m,andy,andonly
exceeded byt.
53.The TIisasound ofdentalposition,and isso
defined andclassed byallthenative authorities.
The peculiarcharacter ofanZ-sound,asinvolving expulsion atthe
sides ofthetongue along with contact atitstip,isnotnoticed byanyHindu
phonetist.
Itisadisputed questionwhether rand Iwere distinguished from one
another inIndo-European speech ;intheSanskrit, atany rate, they arevery
widely interchangeable, both inroots andinsuffixes: there ishardly aroot
containing an Iwhich does notshow alsoforms with r;words written with
theone letter arefound inother texts,orinother partsofthesametexts,
written with theother. Inthelater periodsofthelanguage they aremore
separated,andthe Ibecomes decidedly more frequent, though always much
rarer than ther(onlyas1to7or8or10).
54.Some oftheVedic texts have another -sound, written
with aslightlydifferent character(itisgivenattheendofthe
alphabet, 5),which issubstituted foralingual d(asalso the
same followed byhforadh]when occurring between twovowels.
Itis,then, doubtless alingual I,onemade bybreach(atthe
sides ofthetongue)ofthelingualinstead ofthedental mute-
closure.
55.The <ETyinSanskrit, asinotherlanguages gene-
rally,stands intheclosestrelationshipwith thevowel^i
(shortorlong);thetwoexchangewith oneanother in
cases innumerable.
And intheVeda(asthemetre shows) an iisvery often toberead
where, inconformity with the rules ofthelater Sanskrit euphony, ayis
written. Thus, the final i-vowel ofaword remains ibefore aninitial vowel;
that ofastem maintains itself unchanged before anending; andanending
ofderivation asya,tya has iinstead ofy.Such cases willbecon-
sidered inmore detaillater,asthey arise. Theconstancy ofthephenomenon
incertain words and classes ofwords shows that thiswasnomerely optional
interchange. Very probably, the Sanskrit yhadeverywhere more ofan
i-character thanbelongstothecorresponding European sound.
56.Theyisbyitsphysical character apalatal utterance;
and itisclassed asapalatal semivowelbytheHinduphonetists.
Itisoneofthemostcommon ofSanskrit sounds.
57.The ofvispronounced asEnglishorFrench v
(German w)bythemodern Hindusexcept whenpre-
|
Summarize page 39: | 60] SEMIVOWELS; SPIRANTS. 19
ceded byaconsonant inthesamesyllable,inwhich case
ithasthesound ofEnglish w;andEuropeanscholars follow
thesamepractice (withorwithout thesameexception).
Byitswhole treatment intheeuphonyofthelanguage,
however, thevstands related toanw-vowelpreciselyasyto
an/-vowel. Itis,then, avonly accordingtotheoriginal
Roman value ofthat letter that istosay, az^-sound inthe
English sense :though (aswas stated above forthey]itmay
wellhave been lessmarkedly separated from uthanEnglish w,
more likeFrench ouinouietc. But, astheoriginal whasin
most European languages been changedtov(English),soalso
inIndia, and thatfrom avery earlytime :thePaninean scheme
andtwo ofthePraticakhyas (VPr. andTPr.) distinctly define
thesound asmade between theupperteeth andthelowerlip
which, ofcourse, identifies itwith theordinary modern v-sound.
Asamatter ofpractice,theusualpronunciation need notbe
seriously objectedto
;yetthestudent should not fail tonote
that therules ofSanskrit euphony andthename of"semivowel"
have noapplication excepttoazp-sound intheEnglish sense :
av-sound(German w)isnosemivowel, butaspirant, standing
onthesame articulatestage with theEnglish M-sounds and
the/.
58.Visclassed asalabial semivowelbytheHindu
phoneticalauthorities. Ithasasomewhatgreater frequency
than they.
IntheVeda, under thesame circumstances asthey(above, 55),visto
beread asw.
59.Spirants. Unjler thenameusman(literally 'heat,
steam, flatm\ which isusually and wellrepresented by
'spirant', some oftheHindu authorities include allthe
remaining sounds ofthealphabet;othersapplytheterm
onlytothethree sibilants andtheaspiration towhich
itwillhere alsoberestricted.
Theterm isnotfound inthePaninean scheme; bydifferent treatises
theguttural andlabialbreathings, these andthevisarga, orallthese andanu-
svara, arecalled usman (seeAPr. i.31note). Theorgans ofutterance are
described asbeing inthepositionofthemute-series towhich each spirant
belongs respectively, butunclosed, orunclosed inthemiddle.
60.TheHs.Ofthethreesibilants, orsurdspirants,
this istheoneofplainest and leastquestioned character:
2*
|
Summarize page 40: | 20II.SYSTEM OFSOUNDS.[60
itistheordinary Europeans--ahissexpelled between
thetongue and theroof ofthemouthdirectly behind the
upperfront teeth.
Itis,then, dental, as itisclassed byalltheHindu
authorities. Itistheoneprimitive Indo-European sibilant.
Notwithstandingthegreatlosses which itsuffers inSanskrit
euphony, byconversion totheother sibilants, tor,tovisarga,
etc.,itisstillveryhighamongtheconsonants intheorder of
frequency,orconsiderably morecommon than both theother
two sibilants together.
61.The Ef*.Astothecharacter ofthissibilant, also,
there isnogroundforrealquestion:itistheoneproduced
inthelingual position,orwith thetipofthetonguere-
verted intothedome ofthepalate.Itis,then, akind of
s^-sound; andbyEuropeanSanskritists itispronounced
asanordinarysh(French ch,Germansch\noattempt
being made(anymore than inthecase oftheotherlingual
sounds :45)togiveititsproper lingual quality.
Itslingual character isshown byitswhole euphonic influence,
and itisdescribed andclassed aslingual byalltheHindu author-
ities(theAPr. adds,i.23, that thetongueinitsutterance
istrough-shaped).Initsaudiblequality,itisasA-sound
rather than as-sound; and,inthe considerablevarietyof
sibilant-utterance, even inthesamecommunity,itmaycoincide
withsomepeople'ssh.Yetthegeneral andnormal shispalatal
(seebelow, 63);andtherefore thesign s,marked inaccordance
with theotherlingual letters,istheonlyunexceptionable trans-
literation fortheHindu character.
Inmodernpronunciation inIndia,sismuch confounded with kh;and
theMSS. areapttoexchange thecharacters. Later grammatical treatises,
too,take note oftherelationship (seeWeber'sPratijfia, p.84).
62.This sibilant(aswasnoticed above, 46,and willbe
moreparticularly explained below, 180fF.)isnooriginal sound,
butaproduct ofthelingualization ofsunder certain euphonic
conditions. Theexceptionsareextremely few(9outof145
noted occurrences :75),andofapurely sporadiccharacter. The
Rig-Veda has(apart fromysah,182) only twelve words which
show asunder other conditions.
The final sofaroot has insome cases attained amore independent
|
Summarize page 41: | 65 SPIRANTS. 21
value, and does notrevert toswhen theeuphonicconditions areremoved,
butshows anomalous forms (225.
63.The 5Tg>This sibilant isbyallthenative author-
ities classed and described aspalatal,nor isthereany-
thinginitshistoryoritseuphonictreatment tocastdoubt
on itscharacter assuch. Itis,then, made with the flat
ofthetongue againsttheforwardpartofthepalatalarch
that istosay,itistheusual andnormal sA-sound. By
Europeanscholars itisvariously pronounced moreoften,
perhaps,assthan assh.
Thetwo s/i-sounds, sandp,aremade inthesame part ofthemouth
(thesprobably rather furtherback), butwith adifferent partofthetongue ;
andthey aredoubtless notmore unlike than,forexample, thetwot-sounds,
written tand t;and itwould benotlesspropertopronounce them both as
oneshthan topronounce thelinguals and dentals alike. Toneglect the
difference ofsand fismuch less tobeapproved. Thevery nearrelationship
ofsand fisattested bytheir euphonic treatment, which istoaconsiderable
extent thesame, andbytheir notinfrequentconfusion bythewriters
ofmanuscripts.
64.Aswasmentioned above(41),ther,likec,comes
from thecorruptionofanoriginal &-sound, byloss ofmute-
contact aswell asforward shift ofthearticulating point.In
virtue ofthis derivation,itsometimes(thoughlessoften thanc)
"reverts" tok thatis,theoriginalkappearsinstead of it
;
while, ontheother hand, asas/j-sound,itistoacertain
extent convertible to s.Inpointoffrequency,itslightly
exceeds the latter.
65.Theremaining spirant,^hyisordinarily pronounced
liketheusual Europeansurdaspirationh.
This isnot,however,itstrue character. Itisdefined byallthenative
authorities asnotasurdelement, butasonant(orelseanutterance inter-
mediate between thetwo) ;and itswhole value intheeuphonyofthelanguage
isthat ofasonant: butwhat isitsprecise value isvery hard tosay. The
Paninean scheme ranks itasguttural,asitdoes alsoa:thismeans nothing.
ThePratic.akhyas bringitintonorelation with thegutturalclass :oneofthem
quotes theopinionofsome authorities that "ithasthesame positionwith
thebeginning ofthefollowing vowel" (TPr. ii.47) which sofaridenti-
fies itwith our h.There isnothing initseuphonicinfluence tomark it
asretaining any trace ofgutturally articulated character. Bysome ofthe
native phonetistsitisidentified with theaspirationofthesonant aspirates
|
Summarize page 42: | 22II.SYSTEM OFSOUNDS.[65
with theelement bywhich,forexample, ghdiffers from g.Thisview
issupported bythederivation ofhfrom theaspirates (next paragraph), by
that ofl+hfrom dh(54), andbythetreatment ofinitial hafter afinal
mute (163).
66.The h,asalready noticed,isnotanoriginal sound,
butcomes innearlyallcases from anoldergTi(forthefew
instances ofitsderivation from dhand bh,seebelow, 223).It
isavastly more frequent sound than theunchanged gh(namely,
as7to1):more frequent, indeed, than any oftheguttural
mutes exceptk. Itappears,likej(219),toinclude initself
two stagesofcorruptionofgh:onecorrespondingwith that of
ktoc,theother with that ofktoc;seebelow, 223, forthe
roots belongingtothetwo classesrespectively.Like theother
sounds ofguttural derivation,itsometimes exhibits "reversion"
toitsoriginal.
67.The :h.orvisarga (visarjamya,asitisuniformly
called bythePrati^akhyas andbyPanini, probablyas'be-
longingtotheend' ofaword), appearstobemerelyasurd
breathing,afinal A-sound(intheEuropeansense ofh),
uttered inthearticulating positionoftheprecedingvowel.
OnePraticakhya (TPr.ii.48)gives just this lastdescriptionofit. It
isbyvarious authorities classed withft,orwith handa:allofthem are
alike sounds inwhose utterance themouth-organs have nodefinite shaping
action.
68.Thevisargaisnotoriginal, butalways onlyasubsti-
tute forfinal sorr,neither ofwhich isallowed tomaintain
itself unchanged.Itisacomparatively recent member ofthe
alphabetic system;theother euphonic changes offinal sand r
have notpassed through visargaasanintermediatestage. And
theHindu authorities areconsiderably discordant with onean-
other astohow farhisanecessary substitute, andhow fara
permitted one, alternative with asibilant, before afollowing
initial surd.
69.Before asurdguttural orlabial, respectively, some of
thenative authoritiespermit, while others require, conversion of
final sorrinto theso-calledjihvamutiya andupadhmariiya spi-
rants. Itmaybefairly questioned, perhaps, whether these two
sounds arenotpure grammatical abstractions, devised(likethe
long /-vowel :23)inorder toround outthealphabettogreater
symmetry. Atany rate, neither printed texts normanuscripts
(exceptintherarest andmostsporadic cases) makeanyaccount
ofthem. Whatever individual charactertheymayhavemust be,
|
Summarize page 43: | 71]ANUSVABA. 23
itwould seem,inthedirection ofthe(German)chand^sounds.
When written atall,theyarewont tobetransliterated by%
and(p.
70.The -anusvara, norw
7isanasal soundlacking
that closure oftheorgans which isrequiredtomake a
nasal mute(36);initsutterance there isnasal resonance
alongwithsomedegreeofopennessofthemouth.
71.There isdiscordance ofopinion bothamong theHindu phonetists
andtheirmodern European successorsrespectingthe real character ofthis
element :hence alittle detail isnecessary here with regardtoitsoccurrence
and their views ofit.
Certain nasals inSanskrit areofservilecharacter, always tobeassi-
milated toafollowing consonant,ofwhatever character thatmay be. Such
arefinalminsentence-combination(213), thepenultimate nasal ofaroot,
and anasal ofincrement(255) ingeneral.Ifone ofthese nasals stands
before acontact-letter ormute,itbecomes anasal mute correspondingtothe
latter thatis,anasal utterance inthesame positionofthemouth-organs
which gives thesucceeding mute.If,ontheother hand, thefollowing con-
sonant does notinvolve acontact (being asemivowel orspirant), thenasal
element isalso without- contact :itisanasal utterance withunclosed mouth-
organs. Thequestion is,now, whether this nasal utterance becomes merely
anasal infection ofthepreceding vowel, turning itinto anasal vowel(as
inFrench on,en,un, etc., byreason ofasimilar loss ofanasal mute);or
whether itisanelement ofmore individualcharacter, having place between
thevowel andtheconsonant; or,once more, whether itissometimes theone
thing andsometimes theother. TheopinionsofthePraticakhyas andPanini
arebriefly asfollows :
TheAtharva-Pratic.akhya holds that theresult iseverywhere anasalized
vowel, except when normisassimilated toafollowingI,inthatcase,the
normbecomes anasal I:thatis,thenasal utterance ismade inthe
^-position, andhasaperceptiblei-character.
The other Praticakhyas teach asimilar conversion into anasal counter-
parttothesemivowel, ornasalsemivowel,before yand Iandv(notbefore
ralso). Inmost oftheother cases where theAtharva-Pratic.akhya acknow-
ledges anasal vowelnamely, before randthespirantstheothers
teach theintervention after thevowel ofadistinct nasal element,called the
anusvara, 'after-tone'.
Ofthenature ofthis nasal afterpiecetothevowel nointelligiblyclear
account isgiven. Itissaid(RPr.)tobeeither vowel orconsonant;itis
declared (RPr., VPr.)tobemade with thenosealone,or(TPr.)tobenasal
likethenasal mutes;itisheld bysome (RPr.)tobethesonant tone of
thenasal mutes;initsformation,asinthat ofvowel andspirant,there is
(RPr.) nocontact. Astoitsquantity,seefarther on.
There are,however,certain cases and classes ofcases where these other
|
Summarize page 44: | 24II.SYSTEM OFSOUNDS.[71
authorities alsoacknowledgeanasal vowel.So,especially, wherever afinal
nistreated (208;asifitwere ns(itshistorically olderform); and alsoin
asmall number ofspecified words. They alsomeiitiou thedoctrine ofnasal
vowel instead ofanusvara asheldbysome (and TPr. isuncertain andincon-
sistent initschoice between theoneandtheother).
InPanini, finally,theprevailingdoctrine isthat ofanusvaraevery-
where; and itiseven allowed inmany cases where thePratic.akhyas pre-
scribe only anasal mute. But anasal semivowel isalsoallowed instead be-
fore asemivowel, and anasal vowel isallowed inthecases (mentioned above)
where some ofthePratic,akhyas requireitbyexception.
Itisevidently afairquestion whether thisdiscordance anduncertainty
oftheHinduphonetistsisowingtoareal difference ofutterance indiffer-
entclasses ofcases andindifferentlocalities,orwhether toadifferent scho-
lastic analysis ofwhat isreally everywhere thesame utterance. Ifanu-
svara isanasal element following thevowel,itcannot well* beanything
buteither aprolongation ofthesame vowel-sound withnasality added, ora
nasalised bitofneutral-vowel sound(inthelattercase, however, thealtering
influence ofan iorw-vowel onafollowingsought tobeprevented, which
isnotthecase: see183).
72.The assimilated nasal element, whether viewed as
nasalized vowel, nasal semivowel, orindependent anusvara, has
thevalue ofsomething added, inmakingaheavy syllable, or
length byposition (79).
ThePraticakhyas (VPr., RPr.) give determinations ofthequantity of
theanusvara combining with ashort andwith alongvowelrespectively to
make along syllable.
73.Two differentsigns,1and-,arefound intheMSS.,
indicating thenasal sound here treated of.Usually theyare
written above thesyllable, andtheretheyseem mostnaturally
toimplyanasal affection ofthevowel ofthesyllable, anasal
(anunasika)vowel. Hence some texts(Sama andYajur Vedas),whentheymean arealanusvara, bring one ofthesignsdown
into theordinary consonant-place ;buttheusageisnotgeneral.Asbetween thetwosigns, someMSS.employ,ortend toemploy,the -where anasalized(anunasika) vowel istoberecognized,andelsewhere the1;and this distinction isconsistently observed
inmany European printed texts; and theformer iscalled the
anunasikasign: but itisvery doubtful whether thetwo arenot
originally andproperly equivalent.
Itis^averycommon custom ofthemanuscriptstowrite
theanusvara-signforanynasalfollowing thevowel ofasyllable,either before another consonant orasfinal(notbefore avowel),without anyreference towhether itistobepronounced asnasal
mute, nasalsemivowel, oranusvara. Someprinted texts follow
thisslovenly andundesirable'habit;butmost write anasalmute
|
Summarize page 45: | 751 TABLE OFALPHABETIC SOUNDS. 25
Son.
Surd
Son.
Surdwhenever itistobepronounced excepting where itisan
assimilated m(213).
Itisconvenient also intransliteration todistinguishthe
assimilated mbyaspecial sign, m,from theanusvara ofmore
independent origin, n;and thismethod willbefollowed inthe
present work.
74.This isthewhole systemofsoundsrecognised bythe
written character; forcertain transitional sounds, more orless
widely recognised inthetheories oftheHinduphonetists,see
below, 230.
75.Thewhole spoken alphabet, then,maybearranged
inthefollowing manner, soastoshow, sofarasispossible
inasingle scheme, therelations andimportantclassifications
ofitsvarious members :
Vowels
Semivowels
Nasals
Anusvara
Aspiration
Visarga
Sibilants
111asp.
1-27
bunasp.
.46
phasp.
|
Summarize page 46: | 2(jII.SYSTEM OFSOUNDS. [75
The figuressetunder the characters givetheaverage
percentageoffrequencyofeach sound, found bycounting the
number oftimes which itoccurred inanaggregateof10,OOC
sounds ofcontinuous text, in10different passages,^of1,000
sounds each,selected from different epochsofthe literature :
namely,twofrom theRig-Veda, onefrom theAtharva-Veda,
twofrom different Brahmanas, andoneeachfromManu, Bha-
gavad-Gita, Qakuntala, Hitopade9a,andVasavadatta*.
III.Quantityofsounds and syllables.
76.TheHindu grammarianstake thepainstodefine
thequantityofaconsonant (withoutdistinction among
consonants ofdifferent classes)ashalfthatofashort vowel.
77.Theyalso define thequantityofalong (dirgha)
vowel ordiphthongastwice thatofashortvowel making
nodistinction inthisrespectbetween thegunaand the
prefab-diphthongs.
78.Besides these twovowel-quantities,theHindus
acknowledgeathird,called pluta (literally 'swimming'),
orprotracted, andhavingthree moras,orthree times the
quantityofashort vowel. Aprotractedvowel ismarked
byafollowing figure3:thus, 5TT$a3.
Theprotractedvowels arepracticallyofrareoccurrence(in
RV., three cases; inAV., fifteen; intheBrahman aliterature,
rather morefrequent). Theyareused incases ofquestioning,
especiallyofabalancing between two alternatives, and also of
callingtoadistance orurgently. Theprotractionisofthe last
syllableinawrord, orinawholephrase;and theprotracted
syllablehasusuallytheacute tone, inaddition toany other
accent thewordmayhave
;sometimes ittakes also anusvara, or
ismade nasal.
Examples are: adhdh svidasi3d updri svid asl3t(RV.),'wasit,forsooth,
below? wasit,forsooth, above?' iddmbhUydS id$3m Hi(AV.), 'saying,isthis
more,oristhat?' dgndSi pdtmvdSh s6mam piba (TS.),'ohAgni! thou with
thyspouse!drink thesoma'.
Adiphthongisprotracted byprolongationofits first ora-element:
thus,etoa3i, otoa3u.
*See J.A.0.S.,vol.X.
|
Summarize page 47: | 82]QUANTITY. 37
The signofprotractionisalsosometimes written astheresult ofac-
centual combination, when so-called kampa occurs: seebelow, 90b.
79.For metricalpurposes, syllables (notvowels)are
distinguished bythegrammariansas'heavy' (guru)or'light'
(laghu). Asyllableisheavyifitsvowel islong,orshort
andfollowed bymore than oneconsonant("long bypo-
sition"). Anusvara andvisargacount asfullconsonants in
makingaheavy syllable. The lastsyllableofapada (pri-
marydivision ofaverse)isreckoned aseither heavyor
light.
The distinction interms between thedifference oflong and short in
vowel-sound andthat ofheavy and light insyllable-construction isvaluable,
andshould beretained..
IV.Accent.
80.Thephenomenaofaccentare,bytheHindu gram-
marians ofallages alike, described andtreated asdepend-
ingonavariation oftone orpitch;ofanydifference of
stress involved, theymake noaccount.
81.Theprimarytones(svara)oraccent-pitchesare
two :ahigher (udatta, 'raised'),oracute;and alower
(anudatta, 'notraised'),orgrave. Athird(calledsvarita :
aterm ofdoubtfulmeaning),isalwaysofsecondary origin,
being (when notenclitic :seebelow, 85)theresult ofactual
combination ofanacute vowel andafollowing grave vowel
intoonesyllable.Itisalsouniformlydefined ascompound
inpitch, aunion ofhigher andlower tone within the
limits ofasingle syllable.Itisthus identical inphysical
character with theGreek andLatin circumflex, andfully
entitled tobecalled bythesame name.
82.Strictly, therefore, there isbutone distinction oftone
intheSanskrit accentual system:theaccentedsyllableisraised
intone above theunaccented
;while then further, incertain
cases ofthefusion ofanaccented andanunaccented element
|
Summarize page 48: | 2II.SYSTEM OFSOUNDS.[82
intoone syllable,thatsyllableretains thecompoundedtone of
both elements.
83.The svarita orcircumflex isonly rarely found ona
pure longvowel ordiphthong,butalmost always onasyllable
inwhich avowel, short orlong,ispreceded byayorvre-
presentinganoriginallyacute torw-vowel.
Intransliteration, inthiswork, theudatta oracute willbe
marked with theordinary signofacute, andthesvarita orcir-
cumflex(asbeingadownward slide ofthevoiceforward)with
what isusuallycalled thegraveaccent :thus, d,acute,yaor
va,circumflex.
84.ThePraticakhyas distinguish andname separatelythecircumflexed
tones arising bydifferent processesofcombination :thus, thecircumflex is
called
a.Ksaipra ('quick'), when anacute iorw-vowel(shortorlong)iscon-
verted intoyorvbefore adissimilar vowel ofgrave tone :thus, vyhpta
from vi-apta, apsvantdrfrom apsuantdr.
b.Jatya ('native')ornitya ('own'), when thesame combination lies
further back, inthemake-upofastem orform, and soisconstant, or
belongs toaword inallcircumstances ofitsoccurrence: thus, kva(from fcwa),
svhr(stiar), nybk (nfak), budhnya (budhnfa), kanyh (fcanla), nadyas (nadf-as),
tanvh(tanU-a).
Thewords ofboth these classes areintheVeda, inthegreat majority
ofcases,toberead with restoration oftheacute vowel asaseparate syllable:
thus, apsu antdr, suar, nadias,etc..Insometexts, part ofthem are
writtencorrespondingly:thus, suvar, tanuva, budhnfya.
c.Praflista, when theacute andgrave vowels areofsuch character that
they arefused into along vowel ordiphthong (128):thus divi 'va(RV.
andAV.), fromdM iva; shdgata (TS.), from su-udyata; nai'vhJ
friiyat
(B.), fromndevdapniyat.
d.Abhinihita, when aninitial grave aisabsorbed byafinal acute e
or6(135): thus, te'bruvan, from teabruvan;t>'bravit, from soabravlt.
85.Butfurther, theHindugrammarians agreeinde-
claringthe(naturally grave) syllable following anacute,
whether inthesame orinanother word,tobesvarita or
circumflex--unless, indeed, itbeitself followed byan
acute orcircumflex; inwhich case itretains itsgravetone.
This iscalledbyEuropean scholars theenclitic ordepend-
entcircumflex.
Thus, intenaand teca,thesyllable naandword caare
regarded andmarked ascircumflex :but intena teand teca
svarthey aregrave.
|
Summarize page 49: | 87]METHODS OFWRITING ACCENT. 29
This seems tomean that thevoice, which isborne upatthehigher
pitchtotheend oftheacutesyllable, does notordinarily droptograve pitch
byaninstantaneous movement, butdescends byamore orlessperceptible
slide inthecourse ofthefollowing syllable. NoHindu authority suggests
thetheory ofamiddle orintermediate tone forthe enclitic, anymore than
fortheindependent circumflex. Forthemostpart,thetwo areidentified
with oneanother,intreatment anddesignation. The enclitic circumflex is-
likewise divided into anumber ofsub-varieties, with different names: they
areoftoo little consequencetobeworthreporting.
86.The essential difference ofthetwokinds ofcircum-
flex isshownclearly enough bythese facts :a.theindependent
circumflex takes theplaceoftheacute astheproper accent of
aword, while the enclitic isthemereshadowfollowing anacute,.
andfollowingitinanother wordpreciselyasinthesameword;
b.theindependentcircumflex maintains itscharacter in all
situations, while the enclitic before afollowing circumflex or
acute loses itscircumflex character, andbecomesgrave ;more-
over, c.inmanyofthesystemsofmarking accent(below, 88),
thetwo arequite differentlyindicated.
87.Theaccentuation ismarked inmanuscripts onlyofthe
older literature :namely,inthedifferent Vedic texts, intwoof
theBrahmanas(TaittirlyaandQatapatha),and intheTaittiriya-
Aranyaka. There areanumber ofmethods ofwriting accent,
more orlessdifferent from oneanother
;theonefound inMSS .of
theRig-Veda, which ismost widely known, andofwhich most of
theothers areonly slight modifications,isasfollows :theacute
syllableisleftunmarked
;thecircumflex, whether independent
orenclitic, hasabriefperpendicularstroke above; andthegrave
next preceding anacute or(independent)circumflex hasabrief
horizontal stroke below. Thus
stcfjtri juhdti; rp^T tanvh; WTkva.
Theintroductory gravestroke below, however, cannot begivenifanacute
syllable isinitial, whence anunmarked syllable atthebeginning ofaword
istobeunderstood asacute;andhencealso,ifseveral grave syllables precede
anacute atthebeginning ofasentence, theymust allalike have thegrave
sign. Thus,
^:mdrah; ftte;cflf^fHkarisyasi ;rT%TfTTtuvijatd.
Allthegrave syllables, however, which follow amarked circumflex are left
unmarked, until theoccurrence ofanother accented syllable causes theone
which precedesittotake thepreparatory stroke below. Thus,
sudfcikasamdrk ;
jV__jbutH^lfctiH^J JNIH sudfftkasamdrg gdvam
|
Summarize page 50: | 30II.SYSTEM OFSOUNDS.[88
88.The other methods itisnotworth while toattempttosetforth.
Theymaybefound illustrated inthedifferent texts, andexplained bythe
editors ofthem. Inpart,their peculiaritiesconsist inother forms orplaces
giventothegrave andcircumflex signs.'Insome methods, theacute isitself
marked, byaslight stroke above. Inseveral,thteindependent circumflex is
distinguishedfrom the enclitic. Themost peculiar systems arethescanty
andimperfectone oftheQatapatha-Brahmana,with asingle sign, written
below;andthehighlyintricate oneoftheSama-Veda,with adozen different
signs,written above.
89.Inthiswork, aseverything giveninthedevanagarichar-
acters isalso givenintransliteration,itwill ingeneral be
unnecessarytomark theaccent exceptinthetransliterated form
;
where, however, thecase isotherwise, there willbeadoptedthe
method* ofmarking onlytheaccentedsyllables,theacute
andtheindependentcircumflex :thelatter bytheusual svarita-
sign,theformer byasmall u(for udatta]above thesyllable:
thus,
*^3\ indra, Mi4dgne,
These being given, everything elsewhich theHindu theory recognises
asdependent onandaccompanying them can readily beunderstood asim-
plied.
90.Thetheory oftheSanskrit accent, ashere given (aconsistent and
intelligible body ofphenomena), hasbeen overlaid bytheHindu theorists,
especiallyofthePraticakhyas, with anumber ofadded features, ofamuch
more questionable character. Thus :
a.Theunmarked grave syllables, following acircumflex(eitheratthe
endofasentence,ortillthenear approach ofanotheracute),aredeclared
tohave thesame high tonewith the(alsounmarked)acute. They arecalled
pracayaorpracita ('accumulated': because liable tooccur inanindefinite
series ofsuccessivesyllables).
b.Thecircumflex, whetherindependentorenclitic,isdeclared tobegin
onahigher pitch thanacute, and todescend toacutepitchinordinary cases :
the'concluding instant ofitbeing brought down tograve pitch, however,in
the case ofanindependent circumflex which isimmediately followed by
another ascent ofthevoice tohigher pitch (inacute orindependentcir-
cumflex).
This lastcase,ofanindependent circumflex followed byacute orcir-
cumflex, receivespeculiar written treatment. IntheRig-Veda method, a
fignre 1or3issetafter thecircumflexedvowel, according asitisshort or
long, andthesigns ofaccent arethusapplied:
*Introduced byBohtlingk, andused inthePetersburg lexicon andelsewhere.
|
Summarize page 51: | 93]ACCENT. 31
:apsv alntdh from apsu antdh;
raybS 'vdnih from rayo avdnih .
The other methods, more orlessakin withthis, need notbegiven.
Inthe scholastic utterance ofsuch asyllable ismade apeculiar quaver
orroulade ofthevoice, which iscalled kampa orvikampana.
C.Panini gives theambiguous name ofeka$ruti ('monotone')tothepra-
cita syllables, andsays nothingoftheupliftingofthecircumflex toahigher
plane: heteaches, however,adepression below thegrave pitchforthemark-
edgrave syllable before acute orcircumflex, callingitsannatara(otherwise
anudattatara).
91.Thesystem ofaccentuation asmarked intheVedic texts hasassum-
edinthe traditional recitation oftheBrahmanic schools apeculiar and
artificial form,inwhich thedesignated syllables, grave and circumflex
(equally,theenclitic andtheindependent circumflex), have acquired acon-
spicuous value, while theundesignated, theacute, hassunk into insigni-
ficance *.
92.TheSanskrit accent taughtinthenative grammars and
represented bytheaccentuated texts isessentiallyasystemof
word-accentonly. Nogeneral attemptismade(anymore than
intheGreeksystem)todefine ormark asentence-accent, the
effect oftheemphasis andmodulation ofthesentence inmo-
difyingtheindependentaccent ofindividual words. Theonly
approachtoitisseen inthetreatment ofvocatives andpersonal
verb-forms.
Avocative isusually without accentexceptatthebeginning
ofasentence :forfurther details, seethechapter onDeclension.
Apersonal verb-form isusually accentless inanindepend-
entclause, except whenstandingatthebeginningoftheclause :
forfurther details, seethechapter onConjugation.
93.Certain other words also are, usuallyoralways, without
accent.
a.Theparticles ca,vd,u,sma, iva, cid, svid, fta,arealways without
accent.
b.Thesame istrue ofcertain pronouns andpronominal stems :md,me,
ndu, na<, tvd, te,i?am, vas, ena-, tva-.
c.The cases ofthepronominal stem aaresometimes accented andsome-
times accentless.
Anaccentless word isnotallowed tostand atthebegin-
ningofasentence :alsonotofapadaorprimarydivision of
averse
;apada is,inallmattersrelatingtoaccentuation, treat-
edlikeanindependentsentence.
*Hang, Wedischer Accent, inAbh. d.Bayr. Akad.,vol.XIII, 1874.
|
Summarize page 52: | 32II.SYSTEM OFSOUNDS.
94.Some words havemore than asingleaccentedsyllable.
Such are :
a.Dual collective compounds:asfndravdrunau.
b.Afewother compounds,inwhich eachmember irregularly retains
itsownaccent :astdnundpat, vdnaspdti, brhaspdti.Inarare case ortwo,
also their further compounds,asbrhaspdtipramitta.
C.Infinitive datives intavdf: asetavaf,
d.Aword naturally barytone, buthaving itsfinal syllable protracted:
seeabove, 78.
e.The particle vdvd (intheBrahmanas).
95.OntheplaceoftheaccentedsyllableinaSans-
kritword there isnorestriction whateverdepending upon
either thenumber orthequantityoftheprecedingor
following syllables.The accent rests where therules of
inflection orderivation orcomposition place it,without
regardtoanythingelse.
Thus, indre, agnau, indrena, agnina, agriintim, bahucyuta,
dnapacyuta, parjdnyajinvita,abhimatisahd}dnabhimlatavarna, abhicas-
ticatana, hiranyavacimattama.
96.Since theaccent ismarkedonlyintheolder litera-
ture, and thestatements ofthegrammarians, with the
deduced rules ofaccentuation, arefarfrombeingsufficient
tosettle allcases, theplaceofthe stress ofvoice fora
considerablepartofthevocabularyisundetermined. Hence
itisageneralhabit withEuropeanscholars topronounce
Sanskrit wordsaccordingtotherules oftheLatin accent.
97.Inthiswork, theaccent ofeachword andform will
ingeneral bemarked, sofarasthere isauthority determiningits
place andcharacter. Wherespecific words andforms arequoted,
theywillonly besofaraccentuated astheyarefound with
accent inaccentuated texts.
|
Summarize page 53: | 103] 33
CHAPTER III.
RULES OFEUPHONIC COMBINATION.
Introductory.
98.THE individual elementscomposingalanguage as
actually used are itswords. These are inpart uninflected
vocables(indeclinables, particles);inthemain, theyarein-
flected forms.
99.The inflected forms areanalysable into inflective en-
dings, ofdeclension orofconjugation, and inflected stems to
which thoseendingsareadded.
100. The inflected stems, again, areforthemostpart
asare also inparttheuninflected wordsanalysableinto
derivativeendingsorsuffixes, androots, towhich, eitherdirectly
orthrough more primary stems, those endingsareadded.
But, notafewstems andparticles areirreducible toroots;and,onthe
otherhand, roots areoften used directly asinflectedstems,indeclension as
well asinconjugation.
101. The roots are, inthecondition ofthelanguageas
itliesbefore us,theultimate attainable elements;toagreat
extent notactually ultimate, but,whereotherwise, theresult of
processes ofdevelopmenttooirregular andobscure tobemade
thesubjectoftreatment inagrammar.
102. The formativeprocesses bywhich both inflectional
forms and derivative stems aremade, bytheaddition ofendings
tobases and toroots, aremoreregular andtransparent inSan-
skrit than inanyother Indo-European language, andthegram-
maticalanalysisofwords into their component elements is
correspondingly complete. Hence itbecame themethod ofthe
native grammarians, andhascontinued tobethatoftheirEuro-
pean successors, toteach thelanguage bypresentingtheendings
andstems androots intheir analysed forms, andlaying down
thewaysinwhich these are tobecombined togethertomake
words. And hence astatement oftheeuphonicrules which
govern thecombination ofelementsoccupiesinSanskrit grammar
amoreprominent andimportant placethan inother grammars.
103. Moreover, theformation ofcompound words, bythe
putting together oftwo ormore stems,isaprocessofvery
exceptional frequencyinSanskrit;and thiskind ofcombination
alsohas itsowneuphonicrules. And once more, intheform
Whitney, Grammar. 3
|
Summarize page 54: | 34III-EUPHONIC COMBINATION.[103
inwhich thelanguageishanded down tousbythe litera-
ture, thewords composingasentence orparagraphareadapted
toandcombined with each other bynearlythesame ruleswhich
governthemakingofcompounds,sothat itisimpossibleto
takeapartandunderstand thesimplestsentence inSanskrit
without understandingthose rules. Hence also agreatly added
degreeofpractical importance belongingtothesubjectof
euphoniccombination.
This euphonic interdependenceofthewords ofasentence, which is
unknown toanyother languageinanything like thesame degree, isshown
tobeatleast inconsiderable measure artificial, implying anerection into
necessary and invariable rules ofwhat intheliving language were only
optional practices, bytheevidence oftheolder dialect oftheVedas andthe
younger Prakritic dialects, inboth ofwhich these rules(especially asregards
hiatus: 113) arevery often violated.
104.Wehave, therefore, inthe firstplacetoconsider the
euphonic principles andlawswhich govern thecombination of
theelements ofwords(andtheelements ofthesentence);and
then afterward totakeupthesubjectofinflection, under the
twoheads ofdeclension andconjugation:towhich willsucceed
some account oftheclasses ofuninnected words.
105. The formation ofconjugatioixal stems(tenseand
mode-stems, etc.)willbetaught,asisusual, inconnection
with theprocesses ofconjugationalinflection
;thatofuninflected
words, inconnection with thevarious classes ofthose words.
Butthegeneral subjectofderivation, ortheformation ofde-
clinablestems, willbetakenupbyitself later forabriefpre-
sentation;and itwillbefollowed byanaccount oftheformation
ofcompound stems.
Although, namely, thegeneral plan ofthis series ofgrammars excludes
thesubject ofderivation, yet, because ofthecomparative simplicity and
regularity oftheprincipal processes ofderivation inSanskrit, andtheimport-
ance tothestudent ofaccustoming himself from thebeginning totrace those
processes, inconnection with theanalysis ofderived forms, back totheroot,
anexception willbemade inregard tothesubject inthepresent work.
106.Weassume,then, forthepurposesofthepresent
chapter, theexistence ofthematerial ofthelanguageina
grammatically analysed condition, intheform ofroots, stems,
andendings.
107.What istobetaken astheproper form ofaroot or
stem isnotinallcases clear. Verymanyofboth classes showmapart oftheir derivatives astronger and inapartaweaker
form(260). Thisis,inmostcases, theonly difficulty affecting
|
Summarize page 55: | 108]INTRODUCTORY. 35
stems whether, forexample, weshall speakofderivatives in
mat orinmant, ofcomparatives inyasorinyam,ofaperfect
participleinvatorinvaiis orinus.TheHindu grammarians
usually givetheweaker form asthenormal one,andderive the
other from itbyastrengthening change ;someEuropean author-
ities adopttheoneform andsome theother :thequestionisan
unessential one, givingrise tonopractical difficulty.
108.Asregardstheroots, thedifficultyisgreater, partly
because complicatedwith otherquestions, arising frompractices
oftheHindu grammarians, which have beenmore orlesswidely
followed bytheir Europeansuccessors. Thus :
a.More than half ofthewhole number ofroots given bytheHindu
authorities (which areover2000) have never been found actually used in
theliterature; andalthough some ofthese may yetcome tolight, ormay
have existed without finding their way intoany ofthepreserved literary
documents,itiscertain thatmost arefictitious, made inpartfortheex-
planation ofwords claimed tobetheirderivatives, andinpartforother and
perhaps unexplainable reasons. Ofthe roots unauthenticated bytraceable
usenoaccount willbemade inthisgrammar or,ifatallconsidered,
theywillbecarefully distinguished from theauthenticated.
b.Those roots ofwhich theinitial nand sareregularly converted to
nand *after certainprefixesarebytheHindu grammarians given asbe-
ginning withnands:noEuropean authority follows thisexample.
c.Anumber ofroots endinginawhich isirregularly treated inthe
inflection ofthepresent-systemarewritten intheHindu lists withdiph-
thongseoratoro;and so,after thisexample, bymany Western scholars.
Here theywillberegarded asa-roots :compare below, 251. The oofsuch
roots, especially,ispurely arbitrary ;noforms made from theroot justify it.
d.Theroots showing interchangeably r,ar,and irand Irorurandur
forms arewritten bytheHindus withr,orwithf,orwith both. Here also
thefisarbitrary and indefensible. Asbetween randar,even thelatest
European authorities areatvariance, and itmaybeleft tofurther research to
settle whether theoneortheother isalone worthy tobeaccepted. Here (mainly
asamatter ofconvenience :compare below, 237)ther-forms willbeused.
e.Intheother cases ofroots showing astronger and aweakerform,
choice isingreat measure amatter ofminor consequence unless further
research andthesettlement ofpending phonetic questions shall show that
theone ortheother isdecidedly thetruer andmoreoriginal. From the
pointofview oftheSanskrit alone, thequestionisoften impossibleto
determine.
f.TheHindusclassify assimpleroots anumber ofderived stems :
reduplicated ones,asdidhi, jagr, daridra;present-stems,asurnu;and
denominative stems, asavadhir, kumar, sabhaj, mantr, santv, arth, andthe
like. These areinEuropean works generally reduced totheir true value.
g.But itisimpossible todraw any definite linebetween these cases
|
Summarize page 56: | 36III.EUPHONIC COMBINATION.[108
andothers inwhich root-forms evidently ofsecondary origin have attained a
degreeofindependentvalue inthelanguage which almost orquite entitles
them torank asindividual roots. Even theweak andstrong forms ofthe
same root asvadandvand,citandcint,mahandmahh mayhave
such adifference ofusethatthey count astwo;oradifference ofinflection
combined with adifference ofmeaning inaroothasthesame effect as
invrvrnoti andvrvrnlte, inhajahati andhajihite; oranevident present-
stem becomes aseparateroot asjinvandpinv. Not afew roots occur
inmore orless clearly related groups, themembers ofwhich areofvarious
degreesofindependence. Thus, aconsiderable class ofrootsshow anadded a;
andsuch asmna anddhma arereckoned onlyasside-forms ofmanand
dham; whileJra,pra, pya, psa,and others, presumably made inthesame
manner, figureasseparate from their probable originals. Many final con-
sonants ofroots have thevalue of"root-determinatives",orelements of
obscure orunknown origin added tosimpler forms. Aclass ofderivative
roots show signs ofreduplication,ascaks, jaks, dudh; orofadesiderative
development,asbhaks andbhiks, ?rus, afes, naks. Yetanother classseem
)ocontain aprepositionfused with aroot,asvyac, ap }andthe later ujh
andvyas.
With most ofsuch cases itisnotthepart ofaSanskrit grammar, but
rather ofageneral Indo-European comparative grammar, todealaccording
totheir historical character. Wemust becontent toacceptasroots what
elements seem tohave onthewhole that value intheexisting condition of
thelanguage; with due recognition ofauthorized differences ofopinion on
many points,aswell asofthe fact that further knowledgewill setmany
things now doubtful inaclearerlight.
PrinciplesofEuphonic Combination.
109.The rules ofcombination areinsomerespects
different, accordingasthey apply
a.totheinternalmake-upofaword, bytheaddition
ofderivative and inflectionalendingstoroots andstems;
b.tothemore externalputting togetherofstems to
makecompound stems, andtheyetlooser andmore acci-
dental collocation ofwords inthesentence.
Hencetheyareusuallydivided into rules ofinternal
combination(orsandhi, -putting together'), and rules of
external combination.
110. Inboth classes ofcases, however, thegeneral princi-
ples ofcombination arethesame and likewise, toagreat
extent, thespecific rules. The differences dependinparton
|
Summarize page 57: | 113]GENERAL PRINCIPLES. 37
theoccurrence ornon-occurrence ofcertain combinations inthe
one class ortheother;inpart, onthedifference oftreatment of
thesame sound asfinal ofaroot orofanending,theformer
being much more persistentthan the latter
;inpart, onthe
occurrence inexternal combination ofcertain changes which
areapparently phoneticbutreallyhistorical;and,most frequent
andconspicuousofall,onthefact that(157)vowels andse-
mivowels andnasals exercise asonantizinginfluence inexternal
combination, butnotininternal. Hence, toavoid unnecessary
repetitionaswell astheseparationofwhatreally belongsto-
gether,therules forbothkinds ofcombination willbegiven below
inconnection with one another.
111. Moreover, before case-endings beginningwith bhand
*(namely, bhis, bhyas, bhyam, su},thetreatment ofthe finals
ofstems isingeneralthesame asinthecombinations ofwords
(pada)withoneanother whence those endingsaresometimes
calledjt?ae?a-endings, andthecases theyform areknown aspada-
cases. And withsome ofthesuffixes ofderivation thesame is
thecase.
The importanceofthis distinction issomewhat exaggerated bythe
ordinary statement ofit.Infact, dhistheonly sonant mute initial ofan
ending occurring inconjugation,asbhindeclension;and the difference of
their treatment isinpartowing totheonecoming into collision usually
with thefinal ofarootandtheother ofanending, andinpart tothefact
that dh,asadental,ismore assimilable topalatals andlinguals than bh.
Amore marked andproblematicdistinction ismade between suand the
verbal endings si,sva, etc., especiallyafterpalatal sounds and s.
112. Theleading rules ofinternal combination arethose which areof
highest andmost immediate importancetoabeginner inthelanguage, since
his first task istomaster theprincipal paradigms ofinflection; therules of
external combination may better beleftuntouched until hecomes todealing
with words insentences, ortotranslating. Then, however, they areindis-
pensable, since theproper form ofthewords thatcompose thesentence is
nottobedetermined without them.
Thegeneral principlesofcombination underlyingthe
euphonic rules, anddeterminingtheir classification, maybestated
asfollows :
113.Hiatus. Ahiatus isavoided.
There arebuttwoorthree words inthelanguagewhich
intheiracceptedwritten form exhibit successive vowels forming
differentsyllables:theyaretitau,'sieve'(perhapsfortitasu,
BE,.) andprdilga, 'wagon-pole' (forprayuga ?};and,inRV.,
suuti. For thenotinfrequentinstances ofcompositionand
sentence combination where therecent loss ofsoryorv
leaves apermanent hiatus, seebelow, 132ff., 175b,177.
|
Summarize page 58: | 38III.EUPHONIC COMBINATION.[113
Ithasbeen already noticed that intheVeda, asthemetrical form of
thehymns plainly shows, there isnoavoidance ofhiatus,either asbetween
thestem-finals andendings ofwords, between themembers ofcompounds,
orbetween thewords composingasentence. Incases innumerable, ay
andv(especiallyafter twoconsonants,oralong vowel andconsonant) are
toberead as iandu.But also alongvowel issometimes toberesolved into
two syllables oftenest, aintoa-a :thisresolution issometimes historical,
butordinarily purelymetrical. For details, seebelow. Itiswith regard
tothehiatus thattherules ofthegrammatically regulated classical Sanskrit
aremost demonstrably andconspicuouslydifferent from themore living usages
ofthesacred dialect.
114.Deaspiration. Anaspirate mute isliable to
lose itsaspiration, beingallowed tostand unchanged only
before avowel orsemivowel ornasal.
115. Assimilation. Thegreat bodyofeuphonic
changesinSanskrit,aselsewhere,fallsunder thegeneral
head ofassimilation which takesplace both between
sounds which aresonearlyalike thatthedifference between
them istooinsignificanttobeworthpreserving,andbetween
those which aresodiverse astobepractically incompat-
ible.
116. Inpart,assimilation involves theconversion of
onesound toanother ofthesameseries, without changeof
articulating position ;inpart,itinvolves achangeofposition,
ortransfer toanother series.
117. Ofchanges within theseries, themost frequent and
important occur intheadaptationofsurdandsonant sounds to
oneanother;but thenasals and Ihave also incertain cases
theirspecial assimilative influence. Thus:
a.Inthetwo classes ofnon-nasal mutes andspirants, surdandsonant
arewholly incompatible ;nosurd ofeither class caneitherprecedeorfollow
asonant ofeither.
Amute, surd orsonant,isassimilated bybeing changedtoitscorres-
pondent oftheother kind;ofthespirants, thesurd sistheonlyonehaving
asonantcorrespondent, namely r,towhich itisconvertible inexternal
combination.
b.The nasals aremorefreely combinable: anasalmay eitherprecede
orfollow amute ofeitherkind, orthesonantspirant h;itmay also follow
asurdspirant (sibilant); nonasal, however, everprecedesasibilant inthe
|
Summarize page 59: | 121]GENERAL PRINCIPLES. 39
interior ofaword (itischanged instead toanusvara); and inexternal com-
bination their concurrence isusually avoided byinsertion ofasurdmute.
c.Asemivowel has still lesssonantizing influence;and avowel least
ofall*
:both arefreely preceded and followed bysounds ofevery other
class,intheinterior ofaword.
Before asibilant, however,isfound, ofthesemivowels, only randvery
rarelyI.Moreover,inexternal combination, risoften changed toitssurd
corrspondent5.
But
d.Incomposition andsentence-collocation,initial vowels andsemivowels
andnasals alsorequiretheprecedingfinal tobesonant. And
e.Before anasal andi,the assimilativeprocessissometimes carried
further, bytheconversion ofafinalmute toanasal or Irespectively.
118. Ofconversions involvingachangeofarticulateposi-
tion, themostimportantarethose ofdental sounds tolingual,
and, less often, topalatal. Thus :
a.Thedental sandnare"very frequently converted toandnbythe
assimilating influence ofcontiguous orneighboring lingual sounds :the,
even bysounds namely,iandw-vowels andk which have themselves
nolingual character.
b.Anon-nasal dental mute is(withafewexceptionsinexternal
combination) made lingual when itcomes into collision with alingual sound.
c.Thedental mutes andsibilant aremadepalatal byacontiguous palatal.
But also:
d.Am(not radical) isassimilated toafollowing consonant, ofwhat-
ever kind.
119.Theeuphoniccombinations ofthepalatal mutes, the
palatal sibilant, andtheaspiration,asbeing sounds derived by
phoneticalteration from more original gutturals (42 ff.),are
madepeculiar andcomplicated bytwocircumstances :their rever-
sion toagutturalform(ortheappearanceoftheunaltered
gutturalinstead ofthem^
;andthedifferent treatment of/andh
accordingasthey representoneoranother degreeofalteration
theonetending,likec,more totheguttural reversion, the
othershowing,likec,amore sibilant andlingualcharacter.
120. Thelingualsibilants,also ofderivative character
(fromdental s}:shows asradical final apeculiarandproblematic
mode ofcombination.
121.Extension andabbreviation ofconso-
nant-groups. The native grammariansallow orrequire
certain extensions, byduplicationorinsertion,ofgroupsof
consonants. And, ontheother hand,abbreviation ofcer-
*Inconformity with general phonetic law: seeSievers, Lautphysiologie, p.140.
|
Summarize page 60: | 40III.EUPHONIC COMBINATION.[121
tain other groupsisallowed, andfound often practisedin
themanuscripts.
122.Permitted Finals. Thepermittedoccurrence
ofconsonants attheend ofaword isquite narrowlyre-
stricted. Ingeneral, onlyoneconsonant isallowed after
thelastvowel; and thatmust beneither theaspiration,
norasibilant, norasemivowel (save rarely5J^,noran
aspiratemute, norasonant mute ifnotnasal, norapalatal.
123.Increment andDecrement. Besides these
more orlessregular changes accompanyingthecombination
ofthepartsthatmake upwords, there isanother class of
asomewhat different character, notconsistinginthemutual
adaptationsoftheparts,butinstrengtheningorweakening
changesofthepartsthemselves.
124. Itisimpossibletocarry through aperfectly systematic
arrangement ofthedetailed rules ofeuphonic combination, be-
cause the different varieties ofeuphonic change more orless
overlap and intersect oneanother. The order followed below
willbeasfollows :
1.Rules ofvowel combination, fortheavoidance ofhiatus.
2.Rules astopermittedfinals(sincethese underlie the
farther treatment offinal consonants inexternalcombination).
3.Rules forloss ofaspirationofanaspiratemute.
4.Rules ofsurdandsonant assimilation, including those
forfinal sand r.
5.Rules fortheconversion ofdental sounds tolingual
andpalatal.
6.Rules forthechanges offinal nasals, including those in
which aformer finalfollowing thenasal reappearsincombi-
nation.
7.Rulesregarding thespecial changesofthederivative
sounds thepalatal mutes and sibilant, theaspiration, and
thelingual sibilant.
8.Rules astoextension and abbreviation ofconsonant
groups.
9.Rules forstrengthening andweakening processes.
Everywhere, rules formoresporadic andless classifiable cases
willbegiven inthemostpractically convenient connection;and
theIndex willrender whathelpisneeded toward finding them.
|
Summarize page 61: | 127]VOWEL COMBINATION. 41
Rules ofVowel Combination.
125.Theconcurrence oftwovowels, orofvowel and
diphthong,withoutintervening consonant,isforbidden by
theeuphonyofthelater orclassicallanguage.Itisavoided,
accordingtothecircumstances ofthecase, either byfusion
ofthetwoconcurrent sounds into one.bythereduction of
oneofthem toasemivowel, orbydevelopmentofasemi-
vowel between them.
The texts oftheolder orVedic dialect arewritten according tothe
euphonicrules ofthelater, although inthem(aswaspointed outabove, 113)
thehiatus isreally ofvery frequent occurrence. Hence they arenottobe
read aswritten, butwith almost constant reversal oftheprocessesofvowel
combination which theyhave artificially undergone.
Therules ofvowel combination, asregards both theresult-
ingsound and itsaccent, arenearlythesame ininternal and
inexternal sandhi.
126.Two similarsimplevowels, short orlong,coalesce
andform thecorresponding longvowel: thus, twoa-yowels
(eitherorboth ofthem short orlong) form 5TTa;two^-vow-
els, ^l\twow-vowels. T37w; and, theoretically, twor-vow-
elsform^f,but itisquestionable whether thecase ever
practicallyoccurs. Examplesare:
:saca'prajah (ca-f-aprajah);
ail\a(ati-\-iva) ;
suktam (su-uktam);
cga"sit(raja -\-asit) ;
iadhiqvarah (adhi-~i$varah).
Astheabove examples indicate,itwillbethepractice everywhere in
thiswork,intransliteration(butnotinthedevanagari text), toseparate
independent words;and ifaninitial vowel ofafollowing word hascoalesced
with afinal ofthepreceding,this willbeindicated byanapostrophe
singleifthe initial vowel betheshorter, double ifitbethelonger,ofthe
twodifferent initials which inevery case ofcombination yield thesame result.
127.Ana-vowel combines with afollowing*-vowel to
e/withanw-vowel, toRo;with ^Jr,toSTTar;with
|
Summarize page 62: | 42*
III.EUPHONIC COMBINATION.|127
^Ttheoretically',to3^al;with^eor^ai,to^a*/with
5JT or5tau,to3T#M.Examplesare:
I?RTrajendra (raja-indra);
\\hitopadegah (hita-upadec,ah).;
maharsih (maha-rsih);
va(sa}cvaj;
rajaigvaryam (raja-aigvaryam);
divaukasah (diva-okasah);
jvarausadham (jvara-ausadham).
Insome oftheVedic texts, thevowel riswritten unchangedafter the
a-vowel, which,iflong,isshortened: thus,maharsih instead ofmaharsiti.
Thetwovowels, however,areusually pronouncedasone syllable.
When successive words like indra aihiaretobecombined,the first
combination,toindra,ismadefirst, andtheresult isindre"
'/u(notindrai"
'hi,from indra e'hi).
128.Asregardstheaccent ofthese vowel combinations,itis
tobenoticed that, a.asamatter ofcourse, theunion ofacute
with acuteyields acute, andthatofgravewithgrave yields grave:
thatofcircumflex with circumflex cannot occur;b.acircumflex
with followingacuteyields acute,the final grave element ofthe
former beingraised toacutepitch ;agravewith following acute
does thesame, asnoupwardslide ofthevoice onasyllableis
acknowledgedinthelanguage;but, c.when theformer ofthe
fused elements isacute andthelattergrave, wemight expect
theresulting syllabletobeingeneral circumflex, torepresent
both theoriginal tones. Panini infact allows this accent in
everysuch case; andinasingle accentuated Brahmana text(QB.),
thecircumflex isregularlywritten. Butthelanguage shows, on
thewhole, anindispositiontoallow thecircumflex toreston
eitherlongvowel ordiphthongasitssole basis, andtheacute
element issuffered toraise theother toitsown level ofpitch,
makingthewholesyllable acute. Theonly exceptiontothis,
inmost ofthetexts,isthecombination ofiandt,which be-
comes I:thus, div\'va,from divi iva; intheTaittiriyatexts
alone such acase follows thegeneral rule, while uand u,in-
stead, make ^:thus, shdgata fromsii-udgata.
>129.Thee-vowels, thew-vowels, and ftr,before a
dissimilar vowel oradiphthong,areregularlyconverted
each into itsowncorresponding semivowel, ETyorcfvor
!fr.Examplesare :
|
Summarize page 63: | 132]VOWEL COMBINATION. 43
ityaha(iti-\-aha) ;
madhv iva(madku -f-wa);
duhitrarthe(duhitr-arthe);
stryasya (strl-\-asya);
^tfvadhvdi(vadhu-ai).
But ininternal combination(neverinexternal)the iand
w-vowels arenotseldom changedinstead toiyanduv and
thisespeciallyinmonosyllables,oraftertwoconsonants, where
otherwise agroupofconsonants difficult ofpronunciation would
betheresult. The cases willbenoticed below, inexplaining
inflected forms.
Aradical a-vowel isconverted intoyeven before iinper-
fect tense-inflection :soninyima (nirit -|-ima).
130.Asregardstheaccent here, asinthepreceding
case(128),theonly combinationrequiringnotice isthat ofan
acute iorw-vowel with afollowing grave:theresult iscircum-
flex
;andsuch cases ofcircumflex aremanytimes more frequent
than anyand allothers. Examplesare :
vyusti (vt-usti); ER^ffcT abhyarcati ;
frnadyau (nadi-au);
svuta(su-istaj; H*W tanvas(tanti-as).
Ofasimilar combination ofacute rwith following grave, only asingle
casehasbeen noted inaccented texts :namely, vijndtr etdt(B. xiv. 6.8n
):
theaccentuation isinaccordance with therules for iand u.
131.Ofadiphthong,thefinal iorw-element ischang-
edtoitscorrespondingsemivowel. ^yorof0,before any
vowel ordiphthong:thus,^e(reallyai:28)becomes %may,
and EJTo(that is,au)becomes 5(Sfav;^aibecomes SHITay,
andfaubecomes 35Rav.
^Nochangeofaccent, ofcourse, occurs here; eachoriginal
syllable retains itssyllabic identity, andhence also itsown tone.
Examples canbegiven only forinternalcombination, since inexternal
combination there arefurther changes: seethenext paragraph. Thus,^naya (ne-a); ^TFTnaya (nai-a);
H^bhava(bko-a); >TRbhava(bhau-a).
^132. Inexternalcombination, wehave theimportant
additional rule that thesemivowelresultingfrom thecon-
version ofthe final element ofadiphthongisingeneral
|
Summarize page 64: | 44III.EUPHONIC COMBINATION.[132
dropped;and theresultinghiatus isleftwithout further
change.
133. That istosay.afinal^e(byfarthemost fre-
quent case) becomes simply3fabefore aninitial vowel
(except %a:see135, below), andboth then remain un-
changed;andafinal^ai,inlikemanner, becomes(every-
where) a.Thus,
ttaagatah (te-\-agatah);
nagaraiha(nagare -f-7^/
adadat (tasmai -f-adadat);
uktam(striyai-{~uktam).
The later grammariansallow thei/insuch combinations tobeeither
retained ordropped;buttheuniform practiceoftheMSS.,ofevery age,
inaccordance with thestrict requirementoftheVedic grammars (Pratigakhyas),
istoomit thesemivowel andleave thehiatus.
The persistence ofthehiatus caused bythisomission isaplainindi-
cation ofthecomparativelyrecent loss oftheintervening consonantal sound.
Instances ofthecombination oftheremaining finaland initial arenotunknown,
butthey areofsporadicallyrare occurrence.
/134. The diphthongo(exceptasphoneticalteration of
final as:see175a)isanunusual final, appearing onlyinthe
stemgo(356),inthevoc.sing,ofw-stems, inwords ofwhich
thefinal aiscombined with theparticle u,asatho, andina
fewinterjections. Inthe lasttwo classes itisuncombinable
(below, 138);thevocatives sometimes retain thevandsometimes
lose it(the practices ofdifferent texts aretoodifferent tobe
briefly stated); go(incomposition only)does notlose itsfinal
element, butremains gavorgo.Afinal asbecomes a,with
following hiatus, before anyvowel save a(forwhich, seethe
nextparagraph).
The ofvofSTRavfrom aftauisusuallyretained :thus,
fiic<ef taveva(tau-\-eva) ;
3H"i(e(-*ii4l ubhav indragm (ubhau -\-indragm).
Insometexts, however,itislost before anw-vowel, theaalone
remaining, withhiatus;inatleast onetext(Kathaka),itisdroppedbefore
every vowel. The latergrammarians allow ittobeeither retained ordropped.
135. After final^eor^o.aninitial ?Jadisappears.
Theresulting accent isasiftheawere notdropped,but
rather absorbed into thepreceding diphthong, havingitstone
|
Summarize page 65: | 137JVOWEL COMBINATION. 45
duly representedinthecombination.If,namely,the eorois
grave orcircumflex andtheaacute, theformer becomes acute;
ifthe eoroisacute andtheagrave, theformer becomes cir-
cumflex, asusuallyinthefusion ofanacute and agraveele-
ment. Ifboth areacute orbothgrave, nochange,ofcourse,
isseen intheresult. Examplesare :
te'bruvan(teabruvan);
so'bravit (sdh abravit);
thihsitavyd 'gnih (hinsitavydh agnih);
yaa>indro 'bravit (yddindrah dbramt);
yddrajanyo'bramt (yddrajanydh dbravit).
Astotheuse oftheavagraha sign inthecase ofsuchanelision,see
above, 16. Intransliteration, thereversedapostrophe,orrough breathing,
willbeused inthis \vork torepresentit.
This elision orabsorptionofinitial aafter final eoro,which inthe
later languageistheinvariablerule,isintheVeda only anoccasional
occurrence; and there isnoclose accordance with regard toitbetween the
written andthespoken form oftheVedic texts. IntheAtharvan, forex-
ample, theaisomitted inwriting inabout onethird ofthecases, but is
tobeomitted inreadinginlessthan one fifth(includinganumber inwhich
thewritten text preserves it).SeeAPr. iii.54,note.
Totherules ofvowel combination, asabove stated, there
arecertainexceptions. Some ofthemore isolated ofthese will
benoticed where theycome upintheprocessesofinflection
etc.;afewrequire mention here.
136. Ininternal combination :
a.Theaugmentamakes with the initial vowel ofaroot
thecombinationsat,au,ar(vrddhi-vowels),instead ofe,o,ar
(ywraz-vowels),asrequired by127.
b.The final oofastrengthened stem (238 b)becomes avbefore the
suffix ya(originally za);
C.The final vowel ofastem isoften dropped when asecondary suffix
isadded.
Fortheweakening and loss ofradical vowels, and forcertain insertions,
seebelow, 249ff.,2578.
137. Inexternal combination :
a.The final aoroofapreposition,with initial rofa
root,makes arinstead ofar.
b.The final aofaprepositionbefore roots beginning with eorois
usually omitted.
C.Afinal aincomposition maybecut offbefore otuand ostka.
d.The form uhfrom yvahsometimes makes theheavier(vrddhi)
|
Summarize page 66: | 4(jIII.EUPHONIC COMBINATION.
diphthongalcombination with aprecedinga;thus, prauha. praudhn, akmuhini
(from pra-uha etc.).
138. Certain final vowels, moreover, areuncombinable
(pragrhya) ,ormaintain themselves unchangedbefore any
followingvowel. Thus,
a.Thevowels/,uand easdual endings,both ofdeclen-
sional and ofconjugationalforms. Thus,landliu asate imau;
giriarohatam.
b.Thepronounami,(nom. pi.);andtheVedic pronom-
inalforms asme, yitsme,tve.
c.Afinal omade bycombination ofafinalo-vowel with theparticle u:
thus, atho, mo, no.
d.Afinal IoruofaVedic locative case.
e.Aprotractedfinalvowel(78).
f.The final, oronly, vowel ofaninterjection, asaho, he,a, i,u.
Permitted Finals.
139.Thesounds allowed tooccur asfinals inSanskrit
words, standing bythemselves(notineuphoniccombination
with something following),areclosely limited, and those
which wouldetymologically come tooccupysuch aposition
areoftenvariously altered,ingeneralaccordance with their
treatment inother circumstances, oraresometimes omitted
altogether.
The variety ofconsonants thatwould evercome attheend ofeither
aninflected form oraderivative stem inthelanguageisverysmall :namely,
informs, onlyt(ordj,n,m,s;inderivativestems, only t,d,n,r,s(and,
inafewrarewords, j).Butalmost allconsonants occur asfinals ofroots;
andevery root isliable tobefound, alone oraslastmember ofacompound,
inthecharacter ofadeclined stem.
140. Allthevowel sounds, bothsimpleanddiphthongal,
maybesounded attheendofaword.
Butneitherfnor Iever actually occurs; and risvery rare(only as
neuter sing, ofastem inroror,orasfinal ofsuch astem incomposition).
Thus, indra, civdya, dkari, nadk, d&tu, camu, janayitfj dyne,
civQyai, vnyo, agnau.
141.Ofthenon-nasalmutes, onlythe first ineach
series, thenon-aspirate surd,isallowed; theothers surd
|
Summarize page 67: | 145 PERMITTED FINALS. 47
aspirate,andboth sonants whenever theywould etymo-
logicallyoccur, areconverted into this.
Thus, aynimdfforagnimdth, su/iftforsuhfd,vlrut forvlnid/t.
Inafew roots, when their final sonantaspirate)thus
loses itsaspiration,theoriginalsonantaspirationofthe
initialreappears: compare cjk.below. 147.
Tims, daghbecomes dhak, budhbecomes Wiut, andsoon.
The roots exhibiting thischange arestated below, 155.
There wassome question among theHindu grammariansastowhether
the finalmute istobeestimated asofsurd orofsonantquality;butthe
great weight ofauthority, and theinvariable practice oftheMSS., favor
thesurd.
142.Thepalatals, however, form here(asoften else-
where) anexceptiontotherules fortheother mutes. No
palatalisallowed asfinal. The 3Tcreverts toitsoriginal
efik:thus, oTTfi vak, Jbjc^j^e^anhomuk. The $5"ch(extremely*X *S O^s. L
rare :perhaps onlyintheroot5f^prach) becomes T,t:thus,
"SnTTprat. The sf/either reverts toitsoriginal gutturalor
becomes,inaccordance with itstreatment inother combi-
nations(219): thus, pTERI bhisdk,virat.Thesfi/^ doesnotoccur,
but isbythenative grammariansdeclared convertible to t.
143.Ofthenasals, the *Tmand^nareextremely com-
mon, especiallytheformer(ITmandHsareofallfinal
consonants themostfrequent);the HInisallowed, but is
quiterare;3"nisfound(remainingafter theloss ofafol-
lowing ^1k)inaverysmall number ofwords; 31n
never occurs.
But thefinalmofaroot ischangedton(compare212.
below):thus, dganfromgam,dnanfrom nam.
144.Ofthesemivowels, theFTIalone isanadmitted
final, and itisveryrare.The^ris(likeitsnearest surd
correspondent, Hs:145)changedasfinal tovisarga. Of
ETyand ofvthere isnooccurrence.x17
-v.
145.Ofthesibilants, nonemaystand unaltered atthe
endofaword. TheRs(which ofallfinal consonants
-x \
|
Summarize page 68: | 48III.EUPHONIC COMBINATION.
[
would otherwise bethecommonest) islike^"rchangedtoa
breathing,thevisarga.The$T^?either reverts toitsoriginal
5R,or,insome roots,ischangedto^t(inaccordance
with itschangesininflection andderivation;seebelow, 218).
The *Tsislikewise changedtoJ.
Thechange of to tisofvery rare occurrence:seebelow, 226.
Final radical sissaidbythegrammarianstobechanged tot:thus,
dhvat from dhvas :butnoexampleofthe conversion appears tooccur :
see168.
146.Thecompound^{ksisprescribedtobetreated as
simple*T (notbecomingsfjkby150, below). Butthecase
isarareone.and itsactual treatment intheolder language
irregular.
IntheonlyRV. caseswhere thekshasaquasi-radical character namely
andk from anafcs, anddmyak fromymyakstheconversion istok.Also
offorms ofthes-aorist(seethis aorist below) wehave adhak, asrafc, araife,
etc. for(adhdks-t etc.); but also aprat, dbhrat, ayat (forapraks-t etc.).
AndAV. hasintwo cases srah(i.e.sras), apparentlyforasraks-s, from
ysrj (wronglyreferred byBR. toj/srcms).
Thenumeral a, 'six',isperhaps better toberegarded assafes, with
itskstreated ass,according totheaccepted rule.
147.Theaspiration ^hisnotallowed tomaintain
itself, but(likej^/and $Fg]either reverts toitsoriginal
guttural form, appearingas3Tk,orischangedtoZt
both inaccordance with itstreatment ininflection; seebe-
low, 222. And, also asininflection, theoriginalsonant
aspirationofafewroots(givenat155) reappears when their
final thusbecomesunaspirated. Where the^hisfrom
originalV^dh(223),itbecomes rTt.
148.Thevisarga andanusvara arenowhere etymolog-
ical finals;theformer isonlythesubstitute foranoriginal
finalH^5or^r;the latter occurs asfinalonlysofaras
some latergrammarians allow ittobesubstituted for *Tm-
149.Apart from thevowels, then, theusual finals,
nearlyintheorder oftheirfrequency,are :#,qm,^n,N
|
Summarize page 69: | 152]PERMITTED FINALS. 49
rft,^k,^p,t;those ofonly sporadicoccurrence are
3"n,^l,tITn,-and,bypermitted substitution,-m.
150. Ingeneral, onlyoneconsonant, ofwhatever kind,
isallowed tostand attheendofaword;iftwo ormore
wouldetymologicallyoccur there, thelast isdropped, and
againthelast,and soon,tillonlyoneremains.
Thus, tudants becomes tudant, andthistudan;udanc-s becomes
udank, and thisudan; andachdntst(s-aor.,3dsing.,ofVchand]
isinlikemanner reduced toachan.
But anon-nasal mute,ifradical andnotsuffixal,isre-
tained after r:thus, urkfromurj,vdrkfrom)A>r/,dmart from
ymrj,suhurt from suhdrd. The case isnotacommon one.
For relics offormer doublefinals, preserved bythelater language under
thedisguise ofapparent euphonic combinations, seebelow, 207 ff.
151. Anomalous conversions ofafinalmute toone ofanother class are
occasionally met with. Examples are:
a.Offinal ttofc,inafewwords that have assumed aspecial value
asparticles: thus, jytik, tdjdk (beside tajdt], prthak, rdhak;butalsonow
andthen inaverbal form,assavisak(AV.andVS. Kan.), avisyak (Parask.);
and inthefeminines inkm(asasiknl besideasita). SV. hasonce
prks'iforprtsu.
b.Ofkorjtof,inanisolated exampleortwo,assamydt and dsrt
(TS., K.).
c.InTaittiriya texts, ofthefinal ofanustubh andtristubh toaguttural:
as,anustuk ca,tristiigbhis (Weber, Ind. St.,xiii.109ff.).
d.Ofalabial toadental :inkakud forandbeside kaktibh;insaihsfdbhis
(TS.)fromysrp,andinadbMs, adbhyds, fromaporap(chap. V).Excepting
thefirst,these look like cases ofdissimilation
;yetexamplesofthecom-
bination bbh arenotunknown inthe older language: thus, kakubbhyam,
tristubbhis, anustub bhi.
152. For alltheprocessesofexternal combination that
istosay,incomposition and sentence-collocation astem-
final orword-final isingeneraltoberegardedashaving,not
itsetymological form, butthatgivenitbytherules astoper-
mitted finals. Fromthis, however, are tobeexceptedthe s
and r:thevarious transformations ofthese sounds have nothing
todowith thevisargatowhich asfinals before apause they
have come doubtless atacomparativelyrecent periodof
phonetic historytobereduced. Words willeverywherein
thiswork bewritten with final sorrinstead ofh;and the
rules ofcombination willbestated asforthetwomore original
sounds, andnotforthevisarga.
Whitney, Grammar. 4
|
Summarize page 70: | 50III.EUPHONIC COMBINATION.[153
Deaspiration.
153.Anaspiratemute ischangedtoitscorresponding
non-aspiratebefore another non-nasal mute orbefore asib-
ilant;itstands imaltered onlybefore avowel orsemi-
vowel ornasal.
Such acasecanonly arise ininternal combination, since theprocesses
ofexternal combination presupposethereduction oftheaspirate toanon-
aspiratesurd (152).
Practically, also, therules astochangesofaspirates concern almost only
thesonant aspirates,since thesurd, being oflater development and rarer
occurrence,arehardly ever found insituations that call fortheirapplication.
154. Hence,ifsuch amute istobedoubled,itis
doubled byprefixingitsowncorresponding non-aspirate.
ButintheMSS., bothVedic andlater, anaspirate mute isnotseldom
found written double especially,ifitbeone ofrare occurrence :for
example (BV.), akhkhali, jdjhjhatl.
155. Inafewroots, when afinal sonantaspirate (f
gh,Udh,*Tbh;also^A,asrepresentinganoriginal5Jgh)
thus loses itsaspiration,the initial sonant consonant(JJg
or^"dorSfV)becomesaspirate.
That istosay, theoriginal initial aspirateofsuch roots isrestored,
when itspresence does notinterfere with theeuphonic law,ofcomparatively
recentorigin, which(inSanskrit asinGreek) forbids aroot toboth begin
andendwith anaspirate.
The roots which show thispeculiar changeare :
ingh dagh;
inh(fororiginal gh} dah, dih,duh, druh, drhh, guh; andgrah (in
thelater desiderativejighrksa)]
indhbandh, badh, budh;
inbh dabh(but only inthelater desiderativedhipsa,forwhich the
olderlanguage hasdipsa).
Thesamechange appears when thelaw astofinals causes theloss of
theaspiration attheendoftheroot: seeabove, 141.
Butfromdah, duh,andguharefound intheVeda alsoforms without
therestored initial aspirate: thus, daksnt; aduksat, duduksaetc.; juguksa.
Thesameanalogyisfollowed bydadh, theabbreviated substitute ofthe
present-stem dadha, fromydha,insome oftheforms ofconjugation; thus,
dhatthas fromdadh-+-thas, adhatta fromadadh+ta,etc.
|
Summarize page 71: | 160]ASSIMILATION. 51
Surd andSonant Assimilation.
156.Under thishead, there isespecially oneverymark-
edandimportantdifference between theinternal combi-
nations ofaroot orstem with suffixes andendings, and
theexternal combinations ofstemwith stem incomposition
andofword withword insentence-making: namely,
157. Ininternal combination,the initial vowel orse-
mivowel ornasal ofanendingofderivation orinflection
exercises .noalteringinfluence uponafinalconsonant ofthe
root orstem towhich itisadded.
Tothisrule there areafewexceptions only:thus, arevertedpalatal
sometimes before anasal (216. 4,5) ;dbefore theparticiple-suffix na(161) 5
afinal consonant before maya(161).
Inexternal combination, ontheother hand, aninitial
sonant ofwhateverclass, even avowel orsemivowel or
nasal, requirestheconversion ofafinal surd tosonant.
Ithasbeen pointed outabove (152) that intherules ofexternal com-
bination onlyadmittedfinals, along with sandr,need betaken account of,
allothers being regardedasreduced tothese before combining with initials.
158. Final vowels, nasals, and ^T/arenowhere lia-
bletochangeintheprocessesofsurdandsonant assimi-
lation.
Ther,however, hasacorrespondingsurd ins,towhich it
ischangedinexternal combination under circumstances thatfavor
asurd utterance.
159.With theexceptionsabovestated, the collision
ofsurdandsonant sounds isavoided incombinations
and, regularly andusually, byassimilatingthefinal tothe
followinginitial.
Thus, ininternal combination :dtsi, dtti, atthds, attd(yad
-\-sietc.); qagdhi, cagdhvdm (}/cak -\-dhi etc.);inexternal
combination, dbhudaydm,jyog jiva, sadaqitdyah,tristubdpi;
diggaja, sadahd, brhddbhdnu, abjd.
160.If,however, thefinal sonantaspirateofaroot is
followed by rTtor5Tthofanending,theassimilation isin
4*
|
Summarize page 72: | 52III.EUPHONIC COMBINATION.[160
theother direction: thecombination ismade sonant, and
theaspirationofthefinal(lostaccordingto153, above)is
transferred tothe initial oftheending.
Thus, ghwith torthbecomes gdh;dhwith thesame be-
comes ddh,asbaddhd (ybadh -f-ta),runddhds(^rundh -f-thusor
tas);bhwith thesame becomes bdh, aslaMhd(yiabh-\-ta),
labdhvh, (Vldbh -\-foa).
Moreover, h,asrepresenting original gh,istreated inthe
samemanner :thus, dugdhd, dogdhum from duh andcompare
rudhd andlldhd from ruhandUhetc., 222.
Inthiscombination,asthesonant aspirationisnot lostbuttransferred,
therestoration ofthe initial aspiration (155) does nottakeplace.
Indadh fromydha (155, end),themore normal method isfollowed;the
dhismade surd, andtheinitial aspirated:thus, dhatthas, dhattas. AndRV.
hasdhaktam instead ofdagdham fromydagh.
161. Before anasal inexternal combination, afinal
mutemaybesimply made sonant, oritmaybestill fur-
ther assimilated, being changedtothenasal ofitsown class.
Thus,either Iddndmas ort&nndmas, vagmeorvhnme,
bddmahan orbanmahan, tristub nundm ortristum nundm.
Inpractice,theconversion into anasal isalmost invariably made in
theMSB., as,indeed,itisbythePraticakhyas required andnotpermitted
merely. Even bythegeneral grammariansitisrequiredinthecompound
sdnnavati, andbefore matra, and thesuffix maya (really anoun incom-
position): thus, vanmdya, mrnmdya.
Even ininternal combination,final dofarootbecomes nbefore the
participle-suffix na :thus, bhinnd, sannd, tunnd.
162. BeforeI,afinal tisnotmerely made sonant, but
fully assimilated, becomingI:thus, tdllabhate, uttuptam.
163. Before^h(thecase occursonlyinexternal com-
bination),afinalmute ismade sonant; andthen thehmay
either remainunchangedorbeconverted into thesonant
aspirate correspondingwith theformer: thus, eitherrrf^
tadhiorrrf%taddhi.
Inpractice, the latter method isalmost invariably followed; andthe
grammarians ofthePrati$akhya periodarenearly unanimous inrequiringit.
Thephonetic difference between thetwo isvery slight.
Examplesare:vagghutdh, sdddhota(sat-\-hota),tod-
dhita(tat-\-hita),anustub bhi.
|
Summarize page 73: | 169]FINAL sAND r. 53
Combinations offinalsandr.
164.Theeuphonic changesofHsand^"rmaybest
beconsidered together, because ofthepracticalrelation of
thetwo sounds,incomposition andsentence-collocation,
ascorrespondingsurdandsonant :inahost ofcasesHs
becomes^Trinsituationsrequiringorfavoringtheoccur-
rence ofasonant;and,lessoften, ^~rbecomes Hswhere
asurd isrequired.
Ininternal combination, thetwo arefarlessexchangeable
with oneanother :and this class ofcasesmay bestbetaken up
first.
165. Final rradical orquasi-radical (that is,notbelonging
toanendingofderivation)remains unchanged before bothsurd
andsonant sounds, andeven before suindeclension :thus, pi-
parsi, caturthd, cattirsu, pursu.
166. Final radical sremains before asurd ingeneral, and
usuallybefores,asincassi, cassva, aclssu(butthe last isalso
written ackhsu:172):itishoweverdroppedindsi(forassi :
yas-\-si).Before asonant(that is, bh)indeclension,itis
treated asinexternal combination :thus, acirbhis. Before aso-
nant(that is,dh)inconjugation,itisdropped:thus, cadhi,
adhvdm(butcases like the latter may bebyabbreviation[232]
foraddhvam):inedhi(as-\-dhi),thepreceding vowel isanoma-
louslyaltered.
167. Inaveryfewcases,final radical sbefore sischangedtot(per-
hapsbydissimilation):they are,fromyvas,thefuture vatsydmi and aorist
dvatsam; fromyghas,thedesiderative stem jighatsa.
168. Accordingtothegrammarians, the final sofcertain otherroots,
used asnoun-stems, becomes tattheend oftheword, and before bhand
su:thus, dhvat, dhvadbhis, sradbhyas, sratsu. Butgenuine examplesofsuch
change donotappeartohave beenmetwith inuse.
Sporadiccases ofalike conversion arefound intheVeda :namely,
madbhts andmadbhyds from ma's;usddbhis from usds;svdtavadbhyas from
svdtavas;svdvadbhis etc. (notfound inuse), from svdvas. Butthereality
oftheconversion here isopentograve doubt;itrather seems thesubsti-
tution ofaf-stem foras-stem. Thesame istrue ofthechangeofvans
tovatinthedeclension ofperfect participles (chap. V.).
Inthecompounds ducchuna (dus-funa) andpdrucchepa (parus-cepa),the
final softhe firstmember istreated asifat(203).
For tasapparent endingofthe3dsing,ins-verbs,seechap. VIII.
169.Asthefinal consonant ofderivative stems andofin-
flected forms, both ofdeclension andofconjugation,sisextreme-
|
Summarize page 74: | 54III.EUPHONIC COMBINATION. [169
lyfrequent;and itschangesform asubjectoffirst-rate im-
portanceinSanskrit euphony.Ther,ontheother hand,is
quite rare, beingfound onlyincertain forms ofr-stems and
inafewparticles.
Theeuphonictreatment ofsand ryielding preciselythesame result
after allvowels except aand a,there arecertain forms with regard towhich
itisuncertain whether they end insorr,and opinionsdifferrespecting
them. Such areus(orur}ofthegen.-abl. sing,ofr-stems, andus(orur}
ofthe3dplur.ofverbs.
170. a.TheHs,asalreadynoticed(145), becomes
visargabefore apause.
b.Itisretained unchanged only-when followed by
rTtor%thjthesurdmutes ofitsown class.
c.Before thepalatalandlingualsurdmutes ^cand
^ch,Itand ~&th itisassimilated, becomingthesibilant
ofeither classrespectively, namelySTQorEfs.
d.Before thegutturaland labial surdmutes 3\k
and IrM,Qpand Cfiphitisalsotheoreticallyassimil-
ated, becoming respectivelythejihvamuliyaandupadhma-
niya spirants (60);but inpracticethesebreathingsare
unknown, andtheconversion istovisarga.
Examplesare: tob.tataste,caksus te;toc.tataq ca,ta-
syag chaya;tod.nalah kamam, purusahkhanati;yacah prapa,
vrksahphalavan.
171.The first three ofthese rules arealmost universal;to
the lastthere arenumerousexceptions,thesibilant beingre-
tained(or,by180, converted intos),especiallyincompounds ;
but also, intheVeda, even insentence combination.
IntheVeda, theretention ofthe sibilant incompoundsisthegeneral
rule, theexceptionstowhich aredetailed intheVedic grammars.
Inthelaterlanguage,theretention ismainly determined bytheintimacy
ortheantiquity andfrequency ofthecombination. Thus,the final sibilant
ofaprepositionorawordfilling the office ofaprepositionbefore averbal
root iswont tobepreserved; andthat ofastem before aderivative ofykr,
beforepati, before kalpa andkama, and soon.Examplesarenamaskara,
vacaspati, ayuskama, payaskalpa.
TheVedic retention ofthesibilant insentence-collocation isdetailed in
full inthePraticakhyas. The chief classes ofcases are: a.the final ofa
preposition oritslikebefore averbal form;b.ofagenitive before agovern-
ingnoun: asdivas putrdh, iddspade ;c.ofanablative before part: as
|
Summarize page 75: | 175]COMBINATIONS OFFINAL s. 55
himdvatas part; d.ofother less classifiable cases: asdyaus pitd,trisputvd,
ydspdtih, paridhfs pdtati,etc.
172. Before aninitial sibilant STc,^s,HsHsis
^JT> ^.7^ ^
either assimilated, becomingthesame sibilant, oritis
changedintovisarga.
Thenative grammarians areinsomemeasure atvariance(seeAPr. ii.40,
note)astowhich ofthese changes should bemade, and inpartthey allow
either atpleasure. Theusage oftheMSS. isalsodiscordant;theconversion
tovisargaistheprevalent practice, though the sibilant isalso notinfre-
quentlyfound written. Europeaneditors generally write visarga ;butthe
later dictionaries and glossaries make thealphabetic placeofaword the
same asifthesibilant were read instead.
Examplesare :manuh svayamormanussvayam ;indrah curah
orindracqurah.
173. There areoneortwoexceptionstothese rules:
a.Ifthe initial sibilant has asurdmute afterit,the final smaybe
dropped altogether andbysome authorities isrequiredtobesodropped.
Thus, vayavastha orv&yavah stha,' catustanam orcatuhstanam.
With regardtothispoint theusage ofthe different MSS. and editions
isgreatlyatvariance.
b.Before(s,the sisallowed tobecome visarga, instead ofbeing
retained.
174. Before asonant, either vowel orconsonant(ex-
cept^r:see179),Hsischangedtothesonant^r
unless, indeed,itbepreceded by[aor TTa.
Examplesare :devapatir iva,crir iva;manurgacchati, tanur
apsu ;fayoradrstakamah;sarvair gunaih agnermanve.
Forafewcases likeduda$a, duna$a, seebelow, 199.
Theendings 3^qasand^THas(bothofwhich areex-
tremely common)follow rules oftheir own, asfollows :
175. a.Final 5R7as,before anysonant consonant and
before short^a,ischangedtomo andthe 5faafter
itislost.
The resulting accentuation, andthe fact thattheloss ofaisonly oc-
casional intheolder languageoftheVeda, havebeen pointed outabove, 135.
Examplesare :nalonama, brahmanyo vedavit; hantavyo'smi.
b.Final 3^fasbefore anyother vowel than^aloses
itsHs,becoming simple ^a;and thehiatus thus occa-
sioned remains.
|
Summarize page 76: | 5(>III.EUPHONIC COMBINATION.[175
That istosay, asistreated asanoriginal o,orane,would betreated
inthesame situation: see132 4.
Examplesare :vrhadacva uvaca, aditya iva, ndmaiikti.
176. Exceptionstotherules astofinal asare:
a.Thepronounssasandeshas(also syasintheVeda)lose
their sbefore anyconsonant :thus, sadadarca, eshapurusah ;but
sada tusah, so'bravit. Theexclamation Ihos loses itssbefore
allvowels and allsonant consonants.
b.IntheVeda, andmore rarely inthelater language, therule forthe
maintenance ofthehiatus issometimes violated, and theremaining con-
tiguous vowels arecombined intoone :forexample,se'dague,se'morn,
sawJsadhih (forsaidague, saimam, saosadhih).
c.Afewinstances arefound (almostallVedic) ofsapparently changed
torafter a,asafter other vowels :butinnearly every case there istobe
assumed, rather, astem inarbeside that inas,evidences oftheformer
being sometimes found inthekindred languages:thus, informs ofudhas
anddhas (seechap. V.);inamnas (nooccurrence);inbhiivas (second ofthe
trio bhus, bhuvas, svar), exceptinitsoldest occurrences; inavds (once, in
RV.);inusds(voc., andinusarbudh};invddhar andvadhary (RV.);in
vanargu, dnarvi$, vasarhdn, sabardugha, andone ortwoother more doubtful
words; and inaseries ofwords inasingle passage ofTS.andK., viz.
jinvdr, ugrdr, bhimdr, tvesdr, frutdr, bhutdr, and(K.only) putdr.
Inaharpdti (VS.), andvanarsdd andvanarsdd(RV.), weseethesame
change even before asurd consonant.
d.Final asisoncechanged tooinRV. before asurdconsonant: thus,
adopito.
177. Final 5TTCTasbeforeanysonant, whether vowel or
consonant, loses its *Ts,becoming simple5TTa;andthe
hiatus thus occasioned remains.
Themaintenance ofthehiatus inthese cases,asinthat ofoand e
and ai(above, 1334), seems toindicate arecent loss oftheintermediate
sound. Opinions aredivided astowhat thisshould have been. Some of
thenative grammarians assimilate thecase ofastothat ofai,assuming the
conversion toayinboth alike butprobably only asamatter offormal
convenience inrule-making.
178. Final^"r,ingeneral, shows thesameformwhich
Hswould show under thesame conditions. But
-\
a.Originalfinalr,after aora,maintains itself unchang-
edbefore asonant :thus, punar eti,pratar-jitjdhardkmna, dkar
jyotih.
b.Also before asurd,rispreservedinafewVedic compounds:thus,
suarcanas, svarcaksas, yvarpati, svarsd, svbrsati; dhursdd, dhurs&h; purpati.
|
Summarize page 77: | 181]CONVERSION OFsTOs. 57
vdrkaryd, dfirpada, punartta. Insome ofthese, therisoptionally retained
even inthelater language.
c.Ontheother hand,rislost, likes,inone ortwoVedic cases:
aksd fnduh, ahaevd.
179.Adouble risnowhere admitted: ifsuchwould occur,
either byretention ofanoriginalrorbyconversion of*tor,
one risomitted, and thepreceding vowel,ifshort,ismade
longbycompensation.
Insome Vedictexts, however (Yajur-Veda), arbecomes obefore ini-
tialr;thus, svbrohdva.
Conversion of3U to
v1180.The dental sibilant Hsischangedtothelingual
ET5,ifimmediately preceded byanyvowel save Efaand
ma,orbyfikor^"r unless theHsbefinal, orfol-
lowed by^~r.
The assimilating influence ofthepreceding lingual vowels andsemi-
vowel isobvious enough; that of fcandtheother vowels appears tobedue
toasomewhat retracted position ofthetongueinthemouth during their
utterance, causingitstiptoreach theroof ofthemouth more easily ata
pointfurther back than thedental one.
Thegeneral Hindu grammar prescribesthesame change after aIalso;
butthePratic.akhyas give nosuch rule, andphonetic considerations, the I
beingadental sound, aredecidedly against it.Actual cases ofthecom-
bination,ifthey occur atall,areexcessively rare.
Thevowels that cause thealteration ofstosmaybecalled
forbrevity's sake "alterant" vowels.
Asaconsequence ofthis rule,
181. Inthe interior ofaSanskrit word, thedental sis
notusually found after anyvowel save aanda,but, instead
ofit,thelinguals.But
a.Afollowingrpreventstheconversion :thus, usra, tisras,
tamisra. And itisbutseldom made intheforms and derivatives
ofaroot containing anr-element(whetherrorr),whatever
thepositionofthatelement :thus, sisarti, sisrtam, sarisrpd, tistire,
parisrut. To this rule there areafewexceptions,asvisfard,
nistrta, vispardhas, gdvisthira,etc. Inajusranthe final sofa
root ispreserved even immediatelybefore r.
This dissimilating influence ofafollowing r,ascomparedwith thein-
variable assimilating influence ofapreceding r,ispeculiarandproblematical.
b.Therecurrence ofsinsuccessive syllablesissometimes avoided by
|
Summarize page 78: | 58III.EUPHONIC COMBINATION.[181
leadingtheformer sunchanged: thus, sisaksi,but sisaktiydsislsthds, but
ydsislmahi. Similarly,incertain desiderative formations: seebel'-w, 184c.
C.Other cases aresporadic:RV. hastheforms sisice and sisicus(but
sisicatus),andthestems rbisa, kistd, bfsa, busd, bfsaya; asingle rootpis,
with itsderivative pesuka,isfound once in(JR.;forpuns andthe roots-
ninsand hins, seebelow, 183.
182.Ontheother hand(aswaspointedoutabove, 62),
theoccurrence ofsinSanskrit words isnearlylimited tocases
fallingunder this rule :others arerather sporadic anomalies
exceptwhere sistheproductofcorksbefore adental, asin
drastum, caste, tvastar :see218, 221). Thus, wefind:
a.Four roots, kas, Zas, bhas, bha--, ofwhich the last iscommon and
isfound asearly astheBrahmanas.
b.Further, inRV.,dsa, kavdsa, casala, casa, jdldsa, pdsyh, baskdya,
vdsat(forvaksat?), kdsthd(forkaksta, Fick); and, byanomalous alteration
oforiginal s,-sdh(turdsdh, etc.), dsadha, upasttit, andprobably apdsthd and
asthivdnt. Such cases growmorecommon later.
Thenumeralsas,asalready noted,ismore probablysafes.
183. The nasalization ofthealterant vowelor,inother
words,itsbeing followed byanusvara does notpreventits
alteringeffect uponthe sibilant :thus, havihsi, paruhsi. And
thealteration takesplaceinthe initial sofanendingafter the
final sofabase, whether the latter beregardedasalsochanged
tosorasconverted intovisarga:thus,havissu orhavihsu, pa-
russu orparuhsu.
Butthesofpuns (chap. V.)remains unchanged, apparently onaccount of
theretained sense ofitsvalue aspums ,also that ofyhihs,because ofits
value ashins (hinasti etc.); ynins (RV. only)ismore questionable (perhaps
nims, from nam).
184. Theprincipalcases ofalteration ofsininternal com-
bination are :
a.Inendings, inflectional orderivative, beginning with s
su; si,se,sva;sofsibilant-aorist, future, anddesiderative
;
suffixes sna, snu, sya,etc. after afinal alterant vowel or
consonant ofroot orstem, oraunion-vowel :thus, juhosi, $ese,
anaisam, bhavisyami, cucruse, desna, jisnu, viksu, akarsam.
b.The final sofastem before anendingorsuffix :thus,
havisa, Jwvisas, etc., from havis;caksusmant, coctska, manusa,
manusya.
Roots having afinal sibilant(except p)after analterant vowel are
with theexception offictitious onesandpis, nins, hins regarded asend-
ingins,nots;andconcerning thetreatment ofthis sincombination,see
below, 225 ff.
|
Summarize page 79: | 188]CONVERSION OFsTOs. 59
c.The initial sofaroot after areduplication:thus,si-
syade, susvapa, sisasati, coskuyate,sanisvanat.
Exceptedisingeneral aninitial radical sinadesiderative stem, when
thedesiderative-sign "becomes s:thus,sisanisati fromysan,sisanksati from
ysanj.
185. Butthesame changeoccurs also, onaconsiderable
scale,inexternal combination, especiallyincomposition:thus,
Both inverbal forms and inderivatives, the final ior
Mofaprepositionorother likeprefix ordinarily lingualizesthe
initial softheroot towhich itisprefixed ;since suchcombi-
nations areboth ofgreat frequency and ofpeculiar intimacy,
analogouswith those ofroot orstem and affix :thus, abhisttc,
pratisthh, ntsikta, visita; anusvadhdm, suseka.
The principal exceptionsareinaccordance with theprinciples already
laiddown :namely, when theroot contains anr-element, andwhen arecur-
rence ofthesibilant would takeplace. Butthere arealsoothers, ofamore
irregular character; and tliecomplete account ofthetreatment ofinitial
radical safter aprefix would beamatter ofgreat detail, andnotworth
givinghere.
Inafew cases, the initial,usually altered after acertainprefix,
retains thealtered sibilant even after aninterposed aofaugment orreduplic-
ation :thus, abhy astham, pary asasvajat, vyasahanta, nyasadama, abhy
asincan, vyastabhnat; vitasthe,vitasthire.
Much more anomalous istheoccasional alteration ofinitial radical s
after ana-element ofaprefix. Such cases areavastambh(againstnistambh
andprati stambh)arid(accordingtothegrammarians)avasvan.
186. Inother compounds,the final alterant vowel ofthe
firstmember notinfrequently (especiallyintheVeda) lingualizes
the initial softhesecond :forexample, yudhisthira, pitrsvasr,
gosthd, agnistomd, annstubh, trisandhi, divisdd, paramesthin, abhi-
send, pitrsdd, purustutd.
Averyfewcases occur ofthesame alteration after ana-element:thus,
savyasthd, apasthd, upastut; also}/sa/i,when itsfinal, by146, becomes t:
thus, satrdsdt(butsatrasdham).
187. The final softhe firstmember ofacompoundoften
becomes safter analterant vowel :thus, the sofaprepositional
prefix,asnissidhvan, dustdraffordusstdra), aviskrta; and, regu-
larly,asretained instead ofbeing converted tovisargabefore
alabial orguttural mute(171) ,ashavispa, jyotiskft ;tapuspa,
188. Once more, intheVeda, thesamealteration, both ofaninitial
and ofafinals,isnotinfrequent even between thewords composinga
sentence. The cases aredetailed inthePrati^akhya belonging toeachtext,
andareofvery various character. Thus,
a.The initials,especially ofparticles:asusu,Mma,kdmusvft;
|
Summarize page 80: | $0III.EUPHONIC COMBINATION.[188
also ofpronouns:ashisdh; ofverb-forms, especially fromyas:as
hfsthd,divf stha; andinother scattering cases: asustuhi, nusthirdm,
trisadhdsthd.
b.Afinal s,oftenest before pronouns (especially toneless ones): as
(ignis tva, niste,lyus te,cucistvdm, sddhis tdva; butalso inothercases,
andwherever afinal sispreserved,instead ofbeing turned intovisarga,
before agutturalorlabial (171):asirisputvd, dyus krnotu, vdstospdtih,
dyaus pits.
Conversion ofnton.
189.The dental nasal^n,whenimmediately followed
byavowel orby^norJTmorTfyorcfv,isturned in-
tothelingualHInifprecededinthesame wordbythe
lingualsibilant orsemivowel orvowels that istosay,
by*Ts,~$r.OYftrortyf:andthis, notonlyifthe
alteringletter stands immediatelybefore the nasal, but at
whatever distance from thelatter itmaybefound:unless,
indeed, there intervene(aconsonant movingthefront of
thetongue:namely)apalatal (exceptTy) ,alingual,ora
dental.
Wemay thus figure toourselves therationale oftheprocess:inthe
markedproclivityofthelanguage toward lingual utterance, especially ofthe
nasal, thetipofthetongue, when once reverted into theloose lingual position
bytheutterance ofanon-contact lingual element, tends tohang there and
make itsnext nasal contact inthat position: anddoesso,unless thepro-
clivity issatisfied bytheutterance ofalingual mute, ortheorganisthrown
outofadjustment bytheutterance ofanelement which causes ittoassume
adifferent posture. This isnotthecasewith thegutturals orlabials, which
donotmove thefrontpart ofthetongue (and,astheinfluence ofkon
followingsshows,theguttural position favors thesuccession ofalingual):
andtheyistooweakly palatal tointerfere with thealteration (asitsnext
relative,thei-vowel, itselflingualises as).
This isaruleofconstantapplication;and(aswaspointed
outabove)thegreat majorityofoccurrences ofninthelanguage
aretheresult ofit.
190.The rulehasforceespecially
a.Whensuffixes,ofinflection orderivation,areadded toroots or
stemscontaining oneofthealtering sounds :thus, rudrena, rudrdndm, vdrine,
v&rinl, vdrini, datfni, hdrani, dveshani, krindmi, prntiti, ksubhand, ghrnd,
kdrna, vrknd, rugnd, drdvina, isdni, purand, reknas, cdksana, cfkirsamana,
krpamana.
|
Summarize page 81: | 194]CONVERSION OFnTOn, 61
b.When the finalnofaroot orstem comes tobefollowed, ininflec-
tion orderivation, bysuchsounds asallow ittofeeltheeffect ofapreceding
altering cause: thus, from}/ran, rdnanti, rdnyati, rdrana, ardnisus; from
brahman, brdhmand, brdhmani, brdhmand, brahmanya, brdhmanvant.
191. This rule(likethat forthechangeofsto siapplies
strictly and especially when thenasal andthecause ofitsalter-
ation both liewithin the limits ofthesameintegral word;
but(alsolike theother)itisextended, within certainlimits,
tocompoundwords and even, intheVeda, tocontiguous
words inthesentence.
192.Especially,aprepositionorsimilarprefixtoaroot,
ifitcontain rorend ineuphonicrfor s(174), very often
lingualizesthenofaroot orofitsderived stems and forms.
Thus:
a.The initial nofaroot isusually and regularly soaltered, inall
forms and derivatives, after para, pari, pra, nir(for nis), antar, dur(for
dus):thus,para naya,panmyate, prdnudasva;pardnutti, parindma, pranava,
nirnfj, durndfa.
Roots suffering thischange arewritten with initial ninthenative root-
lists. The only exceptionsofimportance arenrt, nabh, nand(very rare),
andnaywhen itspbecomes *(asinprdnasta],
b.The finalnofaroot islingualized insome oftheforms ofanarid
han: thus, prd 'nitz, prdnd, prdhanyate, prahdnana.
c.The class-signs nuandnaarealtered after the roots hiandmi:
thus, pari hinomi, prdminanti(butthelatter notintheVeda).
d.The 1stsing. impv. ending ani issometimes altered :thus, prd
bhavdni.
e.Derivatives bysuffixes containing nsometimes havenbyinfluence
ofapreposition: thus, prayana.
f.Thenoftheprepositionniissometimes altered,likethe initial of
aroot,after another preposition:thus, pranipdta, pranidhi.
193. Incompound words, analtering cause inonemember sometimes
lingualizesanotthenext following member either itsinitial orfinaln,
orninitsinflectional orderivative ending. The exercise ofthealtering
influence canbeseen todependinpartupon thecloseness orfrequencyof
thecompound, oritsintegration bybeing made thebase ofaderivative.
Examplesare :grdmani, trindman, urunasd
,-vrtrahdnam etc. (butvrtraghnd
tc. :195), nrmdnas, drughand; pravdhana, nrpdna, purydna, pitrydna;
svargena, durgdni, usrdydmne, tryangdnam.
194. Finally, intheVeda,an(usually initial)islingualized evenby
analtering sound inanother word. Thetoneless pronounsnasandena- are
oftenest thus affected :thus, pdri nas. prai 'nan, {ndra enam;butalsothe
particle nd,'like': thus, vdrnd;andafewother cases, asvdrndma, punar
nayamasi, agner dvena.
|
Summarize page 82: | 62HI.EUPHONIC COMBINATION.[195
195.Theimmediate combination ofanwith apreceding guttural or
labial seems insome cases tohinder theconversion ton:thus, vrtraghnS
etc.ksubhnati, trpnoti (butinVeda trpnu).
Conversion ofdental mutes tolinguals and palatals.
196.When adental mute comes incontact with, a
lingualorpalatalmute orsibilant, thedental isusually
assimilated, becoming lingualorpalatal respectively.
The cases arethefollowing:
197.Adental surdmute ornasal, orthedentalsibilant,
when immediately preceded byas,iseverywhere converted into
thecorresponding lingual.
Under thisrule,thecombinations at,sth,andsnareverycommon;aa
israrely sowritten, thevisarga being putinstead oftheformer sibilant:
thus, jydtihsuinstead ofjy6tissu.
Those cases inwhich final abecomes tbefore su(226 b)donot, of
course,fallunder this rule.
198. Intheother(comparatively infrequent)cases where
adental ispreceded byalingualininternal combination, the
dental(exceptofsuloc.pi.)becomeslingual. Thus :
a.Adhafter dfrom aofaroot orstem (226 b):asddviddhvam etc.
b.Only averyfew other instances occur: itteand ditto, from}/id;
saddhd(also saddhd andsodha], andsanndm (aaa -(-nam :anomalousgen.
pi.ofaaa) ;trnna(ytrd-\-na). Asmall number ofwords follow thesame
rule inexternal combination: seebelow, 199.
But tadhi (Vedic: ytad -\-dhi] shows loss ofthe final lingualafter assi-
milation ofthedental, andcompensatory lengthening.
Some ofthecases ofabnormal occurrence ofdareexplained inasimi-
larway, asresults ofalingualized andafterward omitted sibilant before d :
thus nlda fromnisda, ypldfrompisd, j/mrdfrommrsd (Zend marezhdd).
Forwordsexhibiting alikechange incomposition, seebelow, 199b.
199. Inexternalcombination,
a.Afinal tisdirected tobeassimilated toaninitial lingual mute :
thus, tat-tlka, taddayate, tat-thalirii, taddhaukate :but thecase never
occurs intheolder language, andvery rarely inthe later. For finalnbe-
fore alingual,see205.
b.An initial dental after afinallingual usually remains
unchanged ;and suoftheloc.pi.follows thesame rule :thus,
satsuyratsu.
Exceptions are: afewcompounds withaaa,'six': namely, sdnnavati,
sannabhi (and one ortwoothers notquotable from theliterature).
|
Summarize page 83: | 204]DENTAL MUTES TOLINGUALS ANDPALATALS. 63
Inafewcompounds, moreover, there appearsalingualized dental, with
compensatory lengthening,after alostlingual sibilant oritsrepresentative:
namely,incertain Vedic compounds with dus :duddbha, duddf, dudhi, du~
ndfa, dundfa; and, inthelanguage ofevery period, certain compounds of
sas, with change ofitsvowel toanalterant quality (asinvodhum and
sodhum: 224b:sddafa, sodhd(alsosaddhd andsaddhd], sodant.
C.Between final tand initials,theinsertion ofatispermitted
or,accordingtosome authorities, required: thus, sdtsahdsrah orsaltsahdsrah.
200. The cases ofassimilation ofadental toacontiguous
palataloccur almost onlyinexternal combination, and before
aninitialpalatal. There isbutonecase ofinternal combina-
tion, namely:
201. A.^ncomingtofollow apalatal mute ininter-
nalcombination isitselfmadepalatal:
Thus, yacnh, (theonlyinstance afterc),yajna, jajne, ajnata,
rkjna, rhjni.
202.Anfinal rTtbefore aninitialpalatal mute isas-
similated toit,becoming ^cbefore rfcor^ch,and sT/
before sf/(^f\jhdoes notoccur).sr^yAfinal ^Tnisassimilated before 5fj,becoming 3Tn.
Allthegrammarians,ofevery period, require this assimilation ofn;
but itismore often neglected, oronly sporadically made, intheMSS.
Fornbefore asurd palatal, seebelow, 208.
203. Before thepalatalsibilant 5Tp,both rTtand^n
areassimilated, becoming respectivelyrjcand oTn;and
then thefollowingST$may be,and inpractice almost
always is,converted to^ch.
Some authorities regard theconversion ofptochaseverywhere oblig-
atory, others asonly optional ;someexcept, peremptorilyoroptionally, a
ffollowed byamute. Andsome require thesame conversion after every
mute savem,reading alsovtpat chutudrl, dnat chuci, anustup charadi, puk
chuci. TheMSS. generally writec/,instead ofccA, asresult ofthecom-
bination oftand p.
Combinations offinal n.
204. Final radical nisassimilated ininternal combination
loafollowing sibilant, becominganusvara.
Thus, vdnsi, vdhsva, vdhsat, mahsydte, tfghahsati.
According tothegrammarians,itistreated before bhandsuindeclen-
|
Summarize page 84: | 4 HI.EUPHONIC COMBINATION.[204
sion asinexternal combination. Butthecases areextremely rare, andRV.
hasrdnsu andvdhsu (theonlyVedic examples).
Final nofaderivative suffix isregularly andusually dropped before a
consonant ininflection andcompositionincomposition, even before a
vowel; and aradical noccasionally follows thesame rule.
Forassimilation ofntoapreceding palatal,see201.
Theremainingcases arethose ofexternal combination.
205. The assimilation ofninexternal combination toa
following sonant palatal and thepalatalsibilant chave been
alreadytreated(2O2, 2O3).
Thenisalso declared tobeassimilated (becoming n)be-
fore asonantlingual (d,dh, n),but the casehardly ever
occurs.
206.Anisalso assimilated toafollowinginitialI,be-
coming (likem :213c)anasal I.
TheMSS. ingeneral attempttowrite thecombination inaccordance
with this rule.
207. Before thelingual anddental sibilants,sands,final
nremains unchanged;butatmayalsobeinserted between
thenasal andthesibilant :thus, thnsdtorthnt sdt;mahhn sdn
ormahhnt sdn.
According tomost ofthegrammarians ofthePraticakhyas (not RPr.),
theinsertion ofthe tinsuch cases isanecessary one. IntheMSS. itis
very frequently made, butnotuniformly.Itisprobablyapurely phonetic
phenomenon,atransition-sound toease thedouble change ofsonant tosurd
and nasal tonon-nasal utterance althoughthenotinfrequent cases in
which finalnstands fororiginal nt(asbharan, abharan, agnimari) mayhave
aided toestablish itasarule. Itsanalogy with theconversion ofnpinto
nch(203)ispalpable.
208. Before thesurdpalatal, lingual, and dental mutes,
there isinserted after finalnasibilant ofeach ofthose class-
esrespectively, before which thenbecomes anusvara :thus,
hqc;ncch;nst;nsth;hst;nsth.
Thisrule, which inthe classical language hasestablished itself inthe
form here given, asaphonetic rule ofunvarying application, really involves
ahistoric survival. The large majorityofcases offinalninthelanguage
(notfarfrom threequarters: seeAPr. ii.26,note)arefororiginal ns;and
theretention ofthe sibilant insuch, cases, when once itshistorical ground
hadbeenforgotten, wasextended byanalogy toallothers.
Practically, theruleapplies only tonbefore candt,since cases in-
volving theother initials occur either notatall,oronly withextreme rarity
(theVeda does notpresent anexampleofany ofthem).IntheVeda, the
insertion isnotalways made, andthedifferent texts have with regard toit
differentusages, which arefully explainedintheir Praticakhyas;ingeneral,
|
Summarize page 85: | 213]COMBINATIONS OFFINAL nANDm. 65
itislessfrequentintheolder texts. When the pdoes notappear between
nandc,thenisassimilated, becoming n(asbefore j:202).
209. Thesame retention oforiginalfinal safter anasal,
andconsequenttreatment of(apparent)final an, in,un,fnas
iftheywere ans, Ins, uns, fns(longnasalized vowel with finals},
shows itself also inother Vedic forms ofcombination, which,
forthesake ofunity, maybebrieflystated heretogether:
a.Final anbecomes an(nasalized a)before afollowing vowel :that is
tosay, ans, with nasal vowel, istreated likeas,with pure vowel(177):
thus, devdn e'hd,upabaddhdn ihd,mahdn asi. This isanextremely com-
mon case, especially inRV. Once ortwice, the sappears ashbefore p:
thus, svdtavdnh pdyuh.
b.Inlikemanner,sistreated after nasalz,u,fasitwould beafter
those vowels when pure, becomingrbefore asonant sound(174), and(much
morerarely)hbefore asurd(170j:thus, rafminr iva,sununr yuvanyuhr ut,
nfnr abhi; nfnh patram.
210. The nasalsn,n,n,occurringasfinals after ashort
vowel, aredoubled before anyinitial vowel :"
thus, pratydnn ud
esi,udydnn aditydh.
This isalso toberegardedasahistoricalsurvival,thesecond nasal
being anassimilation ofanoriginal consonant following the first. Itis
always written intheMSS., although theVedic metre seems toshow* that
theduplication wassometimes omitted.
211. The nasal nandnbefore asibilant areallowed to
insertrespectivelykand t asn(207)inserts t:thus, praty-
dnksomah.
Combinations offinalm.
212. Final radical JTm,ininternal combination,isas-
similated toafollowing mute orspirantinthelatter case,
becoming anusvara ;intheformer, becomingthenasal of
thesame class with themute.
Beforemorv(aswhen final: 143),itischangedton;thus, from
|/gram, dganma, aganmahi, ganvahi, jaganvans (which appeartobetheonly
quotable cases). According tothegrammarians, thesame change ismade in
the inflection ofroot-stems, before bhand su :thus, prafdnbhis, prafdnsu
(from pra$am: pra-f- j/fam). Noderived noun-stem ends inm.
QB.hasonce theanomalous kdmvant, from theparticle kdm.
213. Final fminexternal combination isaservile
sound, beingassimilated toanyfollowingconsonant.Thus :
*Kuhn, inhisBeitrage etc.,iii.125.
Whitney, Grammar.
|
Summarize page 86: | (J5III-EUPHONIC COMBINATION.[213
a.Itremains unchanged onlybefore avowel oralabial
mute.
But also, byananomalous exception, before roftheroot rajinsamrtij
and itsderivatives samrdjnl andsamrajya.
b.Before amute ofanyother class thanlabial,itbecomes
thenasal ofthat class.
c.Before thesemivowelsy, I,vitbecomes, accordingto
theHindu grammarians,anasal semivowel, thenasal counter-
partofeachrespectively (see 71).
d.Beforer,asibilant, orh,itbecomes anusvara(see 71).
TheMSS. and theeditions ingeneral make noattempttodistinguish
thenasal tones arising from theassimilation ofmbefore afollowing semivowel
from that before aspirant.
e.But ifhbeimmediately followed byanother consonant (which canonly
beanasal orsemivowel), themisallowed tobeassimilated tothatfollowing
consonant.
This isbecause thehhasnoposition ofthemouth-organs peculiarto
itself, but isuttered inthepositionofthenext sound. ThePraticakhyas
donottakeanynotice ofthecase.
Ithasbeenpointedoutabove(73)that theassimilated
misgenerally representedintexts bytheanusvara-sign, and
that inthiswork itistransliterated bym(insteadofanasal
mute orw).Also, thatthegeneral grammariansallowmtobe
pronouncedbeforeanyandevery consonant asanusvara.
Thepalatal mutes and sibilant, and h.
214. These sounds show insome situations areversion to
theoriginal gutturals from which theyarederived. The treat-
ment ofjand/j,also,isdifferent, accordingasthey represent
theoneortheother oftwo different degreesofalteration from
theiroriginals.
215. Thepalatals andharetheleast stable ofalphabetic
sounds, undergoing,invirtue oftheir derivative character,alter-
ation inmany cases where other similar sounds areretained.
216. Thus, inderivation, even before vowels, semivowels,
andnasals, reversion toguttural form isbynomeans rare. The
cases arethefollowing:
1.Before aofsuffixa,finalobecomes kinankd, arkd, pakd, vakd,
parka, markd, vfka, prdtika etc.,rc/ca, seka, moka, rokd, fo'fca, tokd, mrokd,
vraskd: final jbecomes gintyagd, bhdga, bhagd, yaga, bhangd, sanga,
varga, marga, mrgd, sarga, vega, bhoga, yugd, yoga, roga: final hbecomes
ghinaghd, maghd, arghd, dlrghd (anddrdghiyas, drdghistha) ,meghd, dogha.
drtigha, mdgha; andindughana.
Inanother series ofderivatives witha,the altered sound appears:
examples areajd, yaja, pucd, foca, vrajd, vevijd, yuja, urjd, doha.
|
Summarize page 87: | 217]COMBINATIONS OFFINAL PALATALS. 67
Before thesuffixes asandana,theguttural only rarely appears:namely,
indnkas, okas, rofcas, ptf/cas, bhdrgas; andinrogana.
2.Before ani-vowel, thealtered sound appears (exceptinthedoubtful
abhogi):e.g.o/f, tuji, rtici, fdci, vivid, rocisnu.
3.Before w,theguttural reappears,asarule(thecases arefew): thus,
anku, vanku, reku, bhrgu.
4.Before n,theexamplesofreversion arefew, exceptofj(becoming g)
before theparticipial ending na :thus, reknas, vagnu (with the final also
madesonant) ;andparticiples bhagnd, rugnd, etc.; andapparently prgna from
Yprc.
5.Beforem(ofwa,man, min),theguttural generally appears: thus,
rukmd, tigmd, fagmd (with sonant change) ;vdkman, sdkman, rukmdn; rgmtn
andvagmm (with sonant change): butdjman, ojman, bhujmdn.
6.Beforer/,thealtered sound isused :thus, pacya, yajya, yajyu, yujya,
bhujyu. Such cases asbhogya andyogya aredoubtless secondary derivatives
frombhoga andyoga.
7.Beforer,thecases arefew,andtheusage apparently divided :thus,
fukrd, ugrd, mrgra;butvdjra andpajrd(?).
8.Before v(ofthesuffixes va,van, vin, etc., andparticipial vans]
thegutturalisregularly preserved:thus, rkvd, pakvd ;fkvan, rikvan, fukvan,
mrgvan, yugvan; vagvfa (withfurther sonant change); ririkvahs, rurukvdhs,
CUfukvdns; fupukvand, pupukvani: also before theunion-vowel iinokivdns
(RV., once). Anexceptionisydjvan.
The reversion ofhinderivation isexhibited only before the suffix a
(andintheparticiple dughana, RV.). The final jwhich isanalogous with
f(219) shows much lessproclivitytoreversion than thatwhichcorresponds
with c.
9.Alikereversion shows itself also tosome extent inconjugational
stem-formation and inflection. Thus, the initial radical becomes guttural
after thereduplicationinthepresentorperfectordesiderative orintensive
stems oftherootsci,cit, ji, hi,han;andhanbecomes ghnontheelision of
a.TheRV.hasvivakmi fromyvac. And before ran etc. of3dpi.mid.we
have gforradical jinasrgran, asrgram, asasrgram (allinRV.).
217. Final rfcofaroot orstem,iffollowed inin-
ternal combination byanyother sound than avowel or
semivowel ornasal, reverts toitsoriginal gutturalvalue,
andshowseverywherethesame form which a5fikwould
show inthesame situation.
Thus, vakti, uvdktha, vaksi, vaksykmi, vagdhi; vagbhis,vaksu;
ulita, ukthd, vaktdr.
And, asfinal cbecomes k(above, 142),thesame ruleapplies
also tocinexternal combination :thus, vakca,vhg api,vanme.
|
Summarize page 88: | 6III.EUPHONIC COMBINATION.[217
Examplesofcremaining unchangedininflection are :ucydte,
riricre, vaci,mumucmdhe.
218. Final 5T?reverts toitsoriginal3\&,ininternal
combination, onlybefore trie"Q^sofaverbal stem orending
(whence, by180, ^fks]-,before rTtand 5Tih,iteverywhere
becomes Efs(whence, by197,^szand"$"s%before VIdh,
Hbh,andHswoftheloc.pi.,aswhen final(145),itre-
gularlybecomes thelingualmute(tor3d).
Thus, dviksata, veksykmi ;vdsti, vista, didestu; dididdhi,
vidbhis.
Butafewroots exhibit thereversion offinal ctokbefore
bhandsu,and alsowhen final(145): theyare die, drc, sprc,
andoptionallynac(always,inV.);andvifhas inV.always
viksu,loc.pl. ;but vit, vidbhis,etc. Examplesarediksamcita,
drgbhis, hrdisprk,ndk.
Examplesofcremaining unchanged before vowels etc. are:
vici, vivicyas, avicran, acnomi, vacmi, ucmdsi.
\$remains irregularly unchanged before pinthecompound vifpdti.
219.Final sf/isinone setofwords treated like rfc,sT -v7
andinanother setlike STf.
Thus, fromyvj':dyukthas, dyukta, yunkte, yukli, yoktra,
yoksyami, yuksu ;yungdhi, dyugdhvam, yugbhis.
Again, frommrjetc.: dmrksat, sraksy&mi; mtirsti, mrstd, srsti,
rastrd;mrddhi, mrddhvdm, radbhis, ratsu, rat.
Totheformer oryuj-cl&ss belong (asshown bytheir quotable forms)
about twenty roots and radical stems: namely, bhaj, saj, tyaj (riot V.), raj
'color', svaj, majj, nij, tij,vij,iandIbhuj, yuj, ruj, vrj, anj, bhanj, fify ;
urj, sraj, bhisdj, dsrj ; also, stems formed with the suffixes ajandy
;383. 5),astrsndj, vanij:andrtvij, though containing therootyaj.
Tothelatter ormrj-class belong onlyabout onethird asmany:namely,
yaj, bhrajj, vraj, raj, bhraj, mrj, srj.
A.considerable number ofj-roots arenotplacedincircumstances toex-
hibit thedistinction; butsuch roots areinpart assignabletoone orthe
other class ontheevidence ofthe related languages*. The distinction
appears, namely, onlywhen thejoccurs asfinal,orisfollowed,either in
inflection orinderivation, byadental mute(t,th,dh), or,innoun-inflec-
tion, bybhorau. Inderivation (above, 216) wefind agsometimes from
themrj-class: thus, mdrga (inapamarga) andsarga; andbefore rofVedic
3dpi.mid. endings, asrgran, asrgram, asasrgram (beside saarjrire) while
from theywj-class occur only yuyujre, ayujran, bubhujrire, with j.
*SeeH&bschmann, inKZ. xxiii. 384 ff.
|
Summarize page 89: | 222]COMBINATIONS OFFINALch, Jcs,h. 69
220. Final chfallsunder therules ofcombination almost
onlyintherootprac/i,inwhich itistreated asifitwere c
(andpracisperhapsitsmoreoriginal form):thus, praksy&mi,
prstd;and also thederivative pracnd. As final and innoun-
inflection(beforebhandsu],itisdirected tobechangedtothe
lingual mute.
Murtd iscalled theparticipleofmurcfc, and agerund murtva isgiven
tothesame root. They (with mUrti] must doubtless come from asimpler
form ofthe root.
Ofjhthere isnooccurrence :thegrammarians declare it
tobetreated like c.
221. Thecompoundksisnotinfrequentasfinal ofaroot
(generallyofdemonstrably secondary origin),orofatense-stem
(s-aorist:seebelow, chap. XI.);and, inthenotveryfrequentcases
ofitsinternal combination,itistreated asifasingle sound,
followingtherules forc:thus, cdkse(caks -j-se),cdksva;caste,
dcasta, dsrastam, dsrsta, tvdstar. Astoitstreatment when final,
see146.We aretaught bythegrammarianstomake such
forms asgordt, gorddbhis, gordtsu (from gordks];andweactually
havesdt, sadbhts, satsu from saks orsas(146, end).
Inthesingle anomalous root vrapc,thecompound fcissaid tofollow
therule forsimple f:thus, vraksydti, dvraksam; vrdstum, vrasta. Its c.
reverts to feinvraska. Itsparticipleisvrknd;ithasaVedic gerund vrktin.
222.The roots infinal^A,likethose in^j,fallinto
two classes, exhibitingasimilardiversityoftreatment, ap-
pearinginthesame kinds ofcombination.
Intheone class, asduh,wehave areversion ofh(asofc]
toaguttural form, and itstreatment asifitwere still its
original gh: thus, ddhuksam, dhoksy&mi; dugdham, dugdhd;ddhok.
d/iuk, dhugbhis,dhuksu.
Intheother class, asruhand sah,wehave agutturalrever-
sion(asofc)onlybefore 5inverb-formation andderivation :thus,
druksat, roksyftmi, saksiyd,saksdni. Asfinal, inexternal combi-
nation, andinnoun-inflection before bhand su,theh(like c]
becomes alingual mute :thus, turasnt, prtanasad ayodhydh,
turasndbhis, turashtsu. Butbefore adental mute(ttth,dh)in
verb-inflection and inderivation,itseuphoniceffect ispeculiarly
complicated:itturns thedental into alingual (aswouldc];but
italsomakes itsonant andaspirate (aswould dh: see160);
and further,itdisappears itself, and thepreceding vowel,if
short,islengthened bywayofcompensation:thus, from ruh
with tacomes rudhd, from lehwith ticomes Udhi, fromguh
with tarcomesgudhdr, frommehwith turncomes medhum, from
|
Summarize page 90: | 70 HI.EUPHONIC COMBINATION.[222
lihwith tasorthascomes lid/ids, from lihwith d/warn comes
lidhvum, etc.
This isasifwehad toassume astransition sound asonantaspirate
lingualsibilant zh,with theeuphoniceffects ofalingual and ofasonant
aspirate (160),itself disappearing under thelawoftheexisting language
which admits nosonant sibilant.
223. The roots ofthetwo classes, asshownbytheirforms
found inuse, are :
ofthe first ordw/j-class :dah, dih, duh, druh, muh, snih
(andthe final ofumih issimilarly treated);
ofthesecond orrttA-claSB :yah, sah,mih, rihorli/i,guh,
ruh, drh/i, trnh, brJi, bank, sprh(?).
Butmuh forms also(notinRV.) theparticiple mudha andagent-noun
mudhdr, aswell asmugdhd andmugdhdr ;anddruhandsnih areallowed by
thegrammarianstodolikewise: such forms asdrudha and sriidha, however,
donotappeartohave beenmetwith inuse.
From roots oftherwft-class wefind also intheVeda theforms gartaruk,
nom. sing., andpranadhrk anddadhfk; andhence puruspfk (theonlyoccur-
rence) does notprove ysprhtobeofthe dwft-class.
Anumber ofother ft-roots arenotproved bytheir occurring forms to
belongtoeither class;they, too, arewithmore orless confidence assigned
totheone ortheother bycomparison with therelated languages*.
Inderivation, before thesuffixa,wehave (216.1) meghd anddlrghd
from roots ofthe rwft-class. Before therverb-endings, wehaveexamples
onlyfromdw/t, with h:thus, duduhre etc.
The rootnahcomes fromoriginal dhinstead ofgh,and itsreversion is
accordingly toadental instead ofaguttural mute: thus, natsydmi, naddhd,
upanddbhis, upanad-yuga.
224. Irregularities ofcombination are :
a.Thevowel risnotlengthened tocompensateforthe loss ofthe
fe-element :thus, drdhd, trdhd, brdhd(theonly cases;andintheVeda their
first syllable hasmetrical value asheavy orlong).
b.Theroots vahandsahchange their vowel to-oinstead oflengthening
it:thus, vodhdm, vodhdm, vodhdr, sddhum. Butfrom sahintheolder
language forms withaaremorefrequent: thus, sadhd, dsadha(also later),
sadhar. The roottrhhchanges thevowel ofitsclass-sign nainto einstead
oflengthening it:thus, trnedhi, trnedhu, atrnet(thegrammarians teach also
trnehmi and trneksi: ifsuch forms areactually inuse,theymust havebeen
made byfalseanalogy with theothers).
These anomalousvowel-changes seem tostand inconnection with the
fact that thecases showing them aretheonly ones where other than an
*SeeHubschmann, asabove.
|
Summarize page 91: | 226 COMBINATIONS OFFINAL s. 71
alterant vowel (180) comes before thelingualisedsibilantrepresentative of
the h.Compare sodafa etc.
Apparently bydissimilation, the final ofvahintheanomalous compound
anadvah ischanged todinstead ofd:see404.
Thelingual sibilant s.
225. Since thelingual sibilant, initsusual andnormal
occurrences,is(182)theproductoflingualizationofsafter
certain alterant sounds, wemight expectfinal radicals,when
(inrarecases)itcomes tostand where ascannot maintain it-
self, torevert toitsoriginal, andbetreated asaswould be
treated under thesame circumstances. That, however, isonly
true inaveryfew instances.
Namely, intheprefix dus (evidently identical withydus);insajus
(adverbially used case-form fromyjus);in(RV.)vives anddvives, fromyvis;
inaiyes (RV.), fromyis;andinatfs, from ciassecondary form ofy$as.
Allthese, exceptthe first two, aremore orlessopentoquestion.
226. Ingeneral,finallingualTsistreated inthesame
manner aspalatal5T$.Thus :
a.Before tand thitremains unchanged, and the latter
areassimilated :e.g. dvistas, dvisthas, dvestum.
This isacommon andperfectly natural combination.
b.Before dh, bh,and su,asalso inexternal combination
(145),itbecomes alingual mute;anddhismadelingualafter
it :thus, pinddhi, viddhi, vivid dhi, dvidflkvam, dvidbhis, dvitsu.
Thesame holds good ofthealtered sofatense-sign:thus, dstoddhvam
(from astos-dhvam}.
Theconversion of*to*(ord)asfinalandbefore &/tandsuisparallel
with thelikeconversion ofp,and ofjandhinthemrjandruh classes of
roots, andperhaps with theoccasional change ofstot(167 8).Itisavery
infrequent case, occurring (save asitmaybeassumed inthecase ofacts)only
once inRV.andonce inAV. (-dvit and-prut), although those texts have
more than40roots with final s;intheBrahmanas, moreover, hasbeen noticed
further only -flit. From pins, RV. hastheirregular form pinak (2dand
3dsing.,forpinas-s andpinas-t}.
c.Before sininternal combination(exceptsuofloc.pi.)
itbecomes k:thus, dveksi, dveksy&mi,adviksam.
This changeisofanomalous phonetic character, and difficult ofex-
planation. Itisalsopractically ofvery rare occurrence. The onlyRV.
examples (apart from pinak, above) areviveksi, fromyvis,andthe desid.
stem ririksa fromyris ;AV.hasonly dviksat anddviksata, andthedesid. stem
fifliksa fromy^lis. Other examples arequotable fromyykra (B. etc.), fis
(QB.), andpis(Khand. Up.); andthey arebytheHindu grammarians pre-
scribed tobeformed from about half-a-dozen other roots.
|
Summarize page 92: | 72III.EUPHONIC COMBINATION.[227
Extension andAbbreviation.
227.Asageneral rule, chisnotallowed bythegramma-
rians tostand inthatform after avowel, but istobedoubled,
becomingcch(whichtheMSS. sometimes writechcfy.
Thevarious authorities disagree with oneanother indetail astothis
duplication. According toPanini, chisdoubled within aword after either
along orashort vowel; and, asInitial, necessarily after ashort and after
the particles dandmd,and optionally everywhere after along. InRV.,
initial chisdoubled after alongvowel of <Jonly, and certain special cases
after ashort vowel areexcepted.Fortherequired usage intheother Vedic
texts,seetheir several Pratoakhyas. TheKathaka writes fororiginal ch
(notchfrom combination oftandnwith c:203) after avowel every-
where cch.
Opinionsare still atvariance astohow farthisduplication hasan
etymological ground, andhow far itisonlyanacknowledgment ofthe fact
thatchmakes aheavy syllable even after ashort vowel (makes "position": 79).
228. Afterr,anyconsonant(saveaspirant before avowel)
isbythegrammarianseither allowed orrequiredtobedoubled
(anaspirate, byprefixingthecorresponding non-aspirate:154).
Some oftheauthorities include, along withr,also horIor-u,ormore
than oneofthem, inthis rule.
Adoubled consonant after risverycommon inMSS. andinscriptions,
asalso innative text-editions andinthe earlier editions prepared byEuro-
pean scholars inlater ones, theduplicationisuniversally omitted.
The firstconsonant ofagroup whether interior, orinitial after
avowel ofapreceding word isbythegrammarians either allowed or
requiredtobedoubled.
This duplicationisallowed byPanini andrequired bythePraticakhyas
inboth, with mention ofauthorities whodenyitaltogether. Forcertain
exceptions,seethePraticakhyas:themeaningofthewhole matter istoo
obscure tojustify thegiving ofdetails here.
230. Other cases ofextension ofconsonant-groups, required
bysome ofthegrammatical authorities, arethefollowing:
a.Between anon-nasal andanasal mute, theinsertion ofso-called
yamas ('twins'),ornasalcounterparts,istaught bythePrati^akhyas (and
assumed inPanini'scommentary): seeAPr. i.99,note.
b.Between handafollowing nasal mute thePraticaktiyas teach the
insertion ofanasal sound called nasikya:seeAPr. i.100, note.
c.Between rand afollowing consonant thePraticakhyas teach the
insertion ofasvardbhakti or'vowel-fragment': seeAPr. i.101 2,note.
Some authorities assume this insertion only before aspirant;theothers
regard itastwice aslong before aspirant asbefore any other consonant
|
Summarize page 93: | 234]ABBREVIATION OFCONSONANT-GROUPS. 73
namely, ahalf oraquarter mora before theformer, aquarter oran
eighth before the latter. One(VPr.) admits itafter Iaswell asr.Itis
variously described asafragment ofthevowel aorofr(or I).
TheRPr. putsasvarabhakti alsobetween asonant consonant and a
following mute orspirant; andAPr. introduces anelement called sphotana
('distinguished) between aguttural andapreceding mute ofanother class.
Foroneortwoother cases ofyetmore doubtful value, seethePrati^akhyas.
231. After anasal, theformer oftwonon-nasal mutes
maybedropped, whether homogeneous onlywith thenasal, orwith
both :thus, yundhiforyungdhi, yundhvdmforyungdhvdm, antdm for
anktdm, chintkm forchintt&m, bhinthd forbhintthd, indhe forinddhe.
This abbreviation, allowed byPanini,isrequired byAPr. (the other
Praticakhyas take nonotice ofit).Itistheusualpractice oftheMSS.,
though thefullgroupisalso often written.
232. Ingeneral, adouble mute(includinganaspirate
which isdoubled bytheprefixionofanon-aspirate)incombi-
nation withanyother consonant isbythemanuscriptswritten
assimple.
That istosay, theordinary usage oftheMSS. makes nodifference
between those groupsinwhich aphonetic duplicationisallowed bytherules
given above (228, 229) andthose inwhich theduplicationisetymological.
Asevery tvafter avowel may alsobeproperlywrittenttv, sodattvff and
tattvd may be,andalmost invariably are, written asdatvd and tatvd. As
kartana isalsoproperly kdrttana, sokarttika (fromkrtti]iswritten askartika.
Soininflection, wehave always, forexample, majfid etc., notmajjnd, from.
majjdn. Even incomposition andsentence-collocation thesame abbreviations
aremade:thus, hrdyotd foxhrddyotd; chindty asyaforchindtty asya. Hence
itisimpossible todetermine bytheevidence ofwritten usage whether we
should regard adhvam oraddhvam (from yas),ddvidhvam orddviddhvam
(from ydvis)asthetrueform ofasecond person plural.
233.Among occasional omissions ofanetymologically justified member
ofaconsonant- group,isofimportance enoughtobehere noticed* that
Asissometimes lost(perhapsafterassimilation)between
twosurdmutes :thus,
a.The initial softhe roots sthaand stabh after thepreposition ut:
thus, utthdtum forutsthdtum, uttabhnoti forutstabhnoti.
b.Thetense-sign8ofthes-aorist(chap. XI.)after afinal consonant
ofaroot before the initial consonant ofanending: thus, achantta (and for
this, by231, achanta]forachantsta, fapta forfapsta, taptam fortapstam.
c.Other instances areonly sporadic:thus, thecompoundrkthd (re-{-stha:
PB.); thecollocations tasmat tute(forstute) andpuroruk tuta- (forstuta-: K.).
Strengthening andWeakeningProcesses.
234. Under thishead, wetakeupfirst thechangesthat
affect vowels, andthen those that affect consonants adding,
|
End of preview.
Dataset Card for Dataset Name
This dataset card aims to be a base template for new datasets. It has been generated using this raw template.
Dataset Details
Dataset Description
- Curated by: Swapwarick
- Funded by [optional]: [More Information Needed]
- Shared by [optional]: [More Information Needed]
- Language(s) (NLP): English and Sanskrit
- License: Creative Commons Attribution Non Commercial Share Alike 3.0
Dataset Sources [optional]
- Repository: [More Information Needed]
- Paper [optional]: [More Information Needed]
- Demo [optional]: [More Information Needed]
[More Information Needed]
- Downloads last month
- 9