ebook img

Syntactic Patterns in Navaho and Huichol PDF

8 Pages·1.083 MB·Indigenous-English
Save to my drive
Quick download
Download
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.

Preview Syntactic Patterns in Navaho and Huichol

Syntactic Patterns in Navaho and Huichol Author(s): Herbert Landar Source: International Journal of American Linguistics, Vol. 33, No. 2 (Apr., 1967), pp. 121-127 Published by: The University of Chicago Press Stable URL: http://www.jstor.org/stable/1263956 Accessed: 18-08-2014 18:54 UTC Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected]. The University of Chicago Press is collaborating with JSTOR to digitize, preserve and extend access to International Journal of American Linguistics. http://www.jstor.org This content downloaded from 86.177.102.88 on Mon, 18 Aug 2014 18:54:22 UTC All use subject to JSTOR Terms and Conditions SYNTACTIC PATTERNS IN NAVAHO AND HUICHOL HERBERT LANDAR CALIFORNIA STATE COLLEGE AT LOS ANGELES 0. Introduction an approach that is not so much a tool as a 1. Navaho patterns well-integrated theory. 2. Huichol patterns While much of Kelley's criticism of Nav- 3. Evaluation aho Syntax is appropriate, I am impressed by my reluctance to dismiss further explora- 0. The striking similarity between tion of what might be styled 'phrase struc- Grimes's Huichol Syntax and my Navaho tural transformations'. The implications of Syntax (reviewed by Mary L. Foster and the theoretical point to which Kelley has Kalon Kelley, IJAL 32.290-91, 291-94) re- objected, as applied to Navaho and in a sults not simply from the rather obvious small way to Huichol, are the major concern influence that Hockett's work has had on of this paper. After an extensive explication both. It results also in no small measure of Navaho syntax in this context, and an from efforts to explore the relationship of attempt to reinterpret some of Grimes's pattern analysis (hierarchies of immediate Huichol data, I shall return to the question constituents) to transformational analysis. of why one might be motivated to persist in Grimes regards transformational theory as studying the various dimensions of phrase a tool in 'the linguist's toolbox' (p. 5). He structural descriptions. describes Huichol immediate constituents The Navaho patterns which will be de- without assigning cut numbers to indicate scribed in detail are those of interrogative hierarchical rankings, and he interprets as sentences (1.1), negative sentences (1.2), transformations certain combinations of con- and miscellaneous sentences, including many stituents which would, in the light of present- with nominalizations (1.3). Finally, forty- day theory, be assigned to the phrase struc- one patterns, a fair sample of a larger num- ture component of a grammar. Some of the ber that could be identified for Navaho, will transformations which he identifies, how- be tabulated (1.4). The discussion of Huichol ever, would not be objected to by propo- patterns will center on phrasal and conjunct nents of generative theory. transformations. In this respect, I think, Grimes's treat- ment is more sophisticated than my own. 1.1. Many interrogative sentences have The implications of my definition of trans- the enclitic is is it? after one or more ICs formation as any function or mapping, which within the sentence. A simple case is yitzaaz Kelley objects to as a deviant definition but it snowed with is: yitzaaz if did it snow? which I think is the commonly accepted one Such a sentence may have ta? query as its in mathematics, the implications of my initial IC: ta? 1 yitzaaz 2 is query: did it definition do in fact meet the requirement snow? With t1P6 t44? last night compare that structures of one type be mapped into ta? 1 tl?66 4 t44? 8 is 2 yitzaaz query: is it those of another type, as will be shown be- last night that it snowed? low. There are restrictions upon such map- Another type of interrogative sentence pings, however, as is now clear to me, which has a? I'd like to know about it after one of make it futile to hope to reconcile an ap- its ICs: with si I compare si a4?w hat about proach which emphasizes the hierarchical me? In sentences with sa?, ta? does not occur organization of immediate constituents with but is may occur. 121 This content downloaded from 86.177.102.88 on Mon, 18 Aug 2014 18:54:22 UTC All use subject to JSTOR Terms and Conditions 122 INTERNATIONAL JOURNAL OF AMERICAN LINGUISTICS VOL. XXXIII A type of interrogative sentence connoting too 1 yA?at?6eh 2 ta it is not good. The ta concern about a possible situation is charac- may be deleted: too yA?at?6h it is not good. terized by the concatenation of the enclitic It must be deleted after ikfi the (particular) s7 with a verbal whose second member is one (who or which... ); cf. lizin it is black, the enclitic ko being thus. Compare yitzaaz too 1 1iiin 2 ta it is not black, too 1 lizin 2 2 ko 1 q4?w hat if it snowed? ikif it is not the black one. It must also be Yet another type of interrogative sentence deleted before the operators -ii la i t44? in includes a determiner with or without is or case I find (that ... ); cf. nahalthin it's rain- Wsa. Thus with haa how?; what? [a deter- ing, teeaail I'll go, nahalthiin 3 la 2 t44? 1 miner], ?Aniltso it is large, we have haa too 2 tee&sal3 ta in case I find that it's rain- ?anftso how large is it?; with haa where ing, I won't go, too 2 nahalthin 4 la 3 t44? 1 (far)? [a determiner], te? from, we have teesai in case I find that it's not raining, haa te?7 from where?; haa 2 t4? 1 s?7 I'd I'll go. like to know, wheref rom? A negative interrogative sentence is A transform which connotes self-question- formed with too not and is is it? [alternant ing involves an enclitic phrase whose second 6s]. Thus yitzaaz is did it snow?, too 2 6s 1 member is ni? in memory. The relatum of yitzaaz (2 ta) didn't it snow? Negative inter- this phrase is composed of a verbal in con- rogative sentences and even some ostensibly struction with the prior relational expression negative sentences in a context of question- taa ?6iya just by itself, in construction ing prompt a 'yes' answer if the hearer with a prior nominal expression which in- agrees that something is not the case. cludes the determiner haa, and the enclitic To indicate lack of experience one uses the la I find. The operators are conveniently adverb t?ah still, yet in construction with shown as follows, with relative rank marked too not as a constitute before an enclitic by lower case roman numerals: haa iii la ii phrase whose second member is ta and t?aaai v ?6iya iii-. ..-i ni? now let's see, how whose relatum is a verbal with an imperfec- (or what) ... ? Compare yit?6 it is (of a tive verb. The operators are t?ah ii too certain description), haa 3 la 2 t?aa 4 ?6iya 3 i-...-ii ta never. Compare yiA4, I eat it, yit?7e 1 ni? now let's see, how did it look (or yiy4 he eats it, 166? fish, ft?e66 was thus, what was it like)? Note that the set of opera- t?ah 2 too 1 166? 3 yiA4a 2 ta I have never tors is mapped into a matrix sentence of any eaten fish, t?ah 3 too 166? 4 yiy44 3 ta 1 given order of complexity. In terms of the ft?66? he had never eaten fish. ordering of rules, priority is given to the Wonder, interest, and concern are con- generation of the matrix. noted when the verbal includes a verb with Another type of interrogative transform naa- again, some more, and the enclitic la? involves a nominal topic which is optionally golly stands in construction with t?ah too a determiner, and a following verbal com- still not, not yet: t?ah iii too ii la? i-.. .-i ta ment. The nominal is a relatum ending with wonder why (something) never (happens) any ta for example, in construction with 94?; the more. Compare naan6it4 he eats it some more, nominal ends, in other words, with the tVah 3 too 2 la? 1 166? 3 naaneita4 2 ta I operators -ii ta i s? I wonder (about ...). wonderw hy he nevere ats fish any rore. Compare h& where? [a determiner], ti at, If t?ah too still not occurs initially in an ?st?i I'm doing or acting (participatingi n an enclitic phrase whose graduated string is event), haa 4 ti 3 ta 2 s? 1 ?6at?i I wonder -ii yQ i t44 when (yeQ as was mentioned where I am? [q after a syllable ending with a front 1.2. Many negative sentences have too vowel plus a consonant], t44? in time past), not before an IC (or matrix) which ends ta is deleted from the string. A construction with the enclitic ta: yaa?t?f6h it is good, with the operators t?ah iv too iii-.. .-ii ye i This content downloaded from 86.177.102.88 on Mon, 18 Aug 2014 18:54:22 UTC All use subject to JSTOR Terms and Conditions NO. 2 SYNTACTIC PATTERNS IN NAVAHO AND HUICHOL 123 t449 before (when still not), however, is fol- who is pounding now, ?atsit ii the one who lowed optionally by a construction which pounds, the smith. includes too not. Transforms with t?ah ia Alone of relative enclitics, i occurs in cer- too iii-. . .-ii yQ i t44? before (when still not) tain negative imperative constructions. Sec- include a verbal with an imperfective verb ond person future, imperfective, progressive and occur either before a verbal with a per- or optative verbs sometimes serve as im- fective verb or an enclitic phrase whose peratives. The second person optative with second member is hit?6? was thus. Compare lako let it be not serves as a negative impera- tg?il plant, litsii? it is red (neuter imperfec- tive. A more pressing negative imperative tive), tg?il 4 liti? 3 eQ 2 t44? 1 niz6ni 2 requires a relative expression consisting of a fit?66 when the plant was red it was pretty; verbal with a second person continuative yiists66h I see it (imperfective), 7yoo1tl4I imperfective, progressive or iterative verb believei t (continuative imperfective), t?ah 5 and i. This relative expression follows t7V too 4 yiists6h 3 Q2 t44 1 too 3 'yootl1444 (alternant t?aa) just and too not. Compare ta 2 Ait66? beforeI saw it I didn't believei t; nitsiits?iin your head, nilmaz you (sg.) are titsin hunger, nileeh you (sg.) become (im- shaking it, nitsiits?iin 2 nilmaz 1 i you're perfective), sinilf? you (sg.) became (per- the one who's shaking your head, t?aa 2 too 1 fective), tipe sheep, la? one, some, sits?44 nitsiits?iin 3 nilmaz 2 i don't shake your away from me, yinilyal you (sg.) ate meat head! A related pattern involves verbals (perfective), titsin 3 sinili? 2 e 1 t44? when which include a first person sg. imperfective you became hungry, t?ah 5 too 4 titgin 5 verb and a following verbal with a perfective nileeh 3 e 2 t44? 1 tipe 3 a? 2 sits?44 3 verb. Compare nisaah I go (move as a single yinilyal beforey ou becameh ungry you ate one entity), ni?ah i I am one who goes, t?aa 2 of my sheep; ?aw6d? baby, haleeh it comes too 1 niMah2 i without my being one who into existence (imperfective),pamA its mother, goes, ts'?ili Chinle, ti at, gee with me, nikhi- taazts4 she has died (perfective), t?ah 5 too 4 honilth4 it started to rain, t?aa 3 too 2 ?awee? 5 haleeh 3 e 2 t44? 1 pama 2 taazts4 tW?iflfi5 ti 4 niah 3 i 1 see 2 nikhihonilth4 beforet he baby was borni ts motherd ied. before I could get to Chinle, I got caught in 1.3. Miscellaneous patterns include many the rain. with relative enclitics. The commonest rela- In patterns where ta follows a negated tive enclitics are ikff, i and ii, all meaning nominalized optative with i, the ta is deleted the one (who or which... ). Verb stems of the before t44? if. Compare t6ya? I wish he type -CVVC alternate with -CVVC before could start going along (-ya? [one] move [opta- ii and -CVC before i and ikii. Whereas tive]), pee with them, t?aa 2 too I pee 4 nominalizations in some languages require t6ya? 3 i 2 ta he is just not the one who could double base transformations, the matter is start going along with them (he has no way of much simpler in Navaho (and, for that going), t?Va 3 too 2 pee 5 t6ya? 4 i 3 t44? 1 matter, Huichol), witness tA?itinistl66h I titeesaal if he has no way of going along with am about to smile, ts?itinistl6h i I am the one them, I'll go. who is about to smile (note that ikii and ii In a warning or stipulation against an undesirable condition, a negated nominalized are not used with the first person sg. verbal imperfective or other verb in i is followed subject); ts?iteeltl6h ikii, ts?iteeltl6oh ii and by ko being thus and t?eiya only: t?ia iv too tP?iteeltl6h i he is the one who is about to iI-.. .-iv i ii ko i t?eiya only on condition smile, and so on for other subjects. In many that (something) will not (be done). Compare cases there is a differentiation of meaning; ?atatUitl66h they [distant] laugh, ?azata?- cf. ?atsit he pounds it, ?atsit ikii the particu- toonil he will carry on a test ceremony,t Vaa 5 lar thing which he pounds, ?atsit i the one too 4 ?atatUitl6h 5 i 3 ko 2 t?6iya 1 ?azata?- This content downloaded from 86.177.102.88 on Mon, 18 Aug 2014 18:54:22 UTC All use subject to JSTOR Terms and Conditions 124 INTERNATIONAL JOURNAL OF AMERICAN LINGUISTICS VOL. XXXIII toonil only if they don't laugh will he carry on ht? 2 nfti 1 yitl6oh 2 ta he obviously isn't a test ceremony.T he verb is not nominalized even feeling cold. The enclitic hanii nay that unless negated. Thus tipe sheep, naakhi generally indicates a condition contrary to two, paa to him, yininil you gave him several fact and is not restricted to constructions objects, tip6 5 naakhi 4 paa 5 yininil 3 ko 2 with too not, such as too 2 hanii 1 yitl6oh 2 t?6iya I ?azata?toonil he will hold a test ta nay that he is not feeling cold; cf. si 2 hanii ceremonyo nly if you give him two sheep. 1 S?6t?i I'm not the one who's doing it. As noted above, too not figures in some The enclitic hti is common in concessive negations but not in others. With optatives transforms with a construction of the pro- one can use t?a just and the adverb kha clitic ?az4 granted and a verbal (or ?a,aan against [khaa before ta]. Compare hooyan nee?, with nee? poor me, and a verbal). hogan, dwelling, k6 ne? back there, inside, Compare i"? 1 tah 2 tiilyot the horse ran off, ha66tziih I wish you would speak out (-tziih tin6 1 t?66 2 neezta the man merely sat, breatheo r speak out, optative), tOaa 2 kha 1 ?ai4 3 l1f?4 tah 5 tiilyot 2 nti 1 tin6 2 t?66 3 hooyan 3 k6 4 ne? 2 hati6tziih see that one neezta even granted that the horse ran off, the doesn't say anything inside the hogan [hat- man merely sat, ?aiaan 4 nee? 3 lii? 4 tah 5 M6tziihI wish that one would speak out], t?a& tiilyot 2 hti 1 tine 2 t766 S neezta even 2 kha 1 hooyan 3 k6 4 ne? 2 ha66tziih see that granted, woe is me, that the horse ran off, the you don't say anything inside the hogan. man merely sat. A further case of too with a nominaliza- Patterns with numerals, with particular tion in i is used in angry or resentful ad- verbs, and in the posing of alternatives monitions; the nominalization follows too might also take up our attention, but I not and y6e emphasis. Compare niilts4 I think that a sufficient sample of what is saw you, niiltsan i I am the one who saw you, implicit in pp. 48-9 of my Navaho Syntax too 2 y6e 1 niiltsan 2 i it's a lucky thing for has been given to clarify what is at issue. you that I didn't see you! 1.4. Transformational operators follow in The idiomatic nature of many patterns tabular form (p. 125). which I have characterized in terms of transformational sets should be apparent by 2. The foregoing patterns cannot be iden- now. The patterns relate to pragmatic situa- tified by scanning a corpus. Informants tions in an interesting way. We will look at must make judgments about them and the some more patterns and then will summarize size and internal variety of the string classes in tabular form those we have seen and some with which they are used. Grimes's Huichol which space does not permit us to give in Syntax, corpus oriented and without IC detail. rankings, offers the added difficulty for one Several transforms with too not as the intent on finding Huichol transformational head of an enclitic phrase connote persuasive sets in the sense in which the term is used negation. In one of these, the enclitic is here that while it contains by far the best ht6? obviously, as everybodyk nows, as is in description of the Huichol verb yet pub- evidence, in another hti even, in another the lished, it confronts one with puzzles of verbal string ht6? nti, in another hanii nay that. morphology. It is possible, nevertheless, to These constructions pattern with an enclitic suggest that certain Huichol operators in phrase whose relatum is a verbal and whose Grimes's treatment of transformations (pp. second member is ta for example. Compare 54-69) are phrase structural. Some are as yitl66h he is feeling cold, too 1 yitl6oh 2 ta simple as tiita i- what? Compare (p. 57) he is not feeling cold, too 2 it6? 1 yitl6oh 2 ta tiita 1 tiyXane 2 ?uuma what is moving he obviously is not feeling cold, too 2 hti 1 over there? Some are of greater complexity. yitl6oh 2 ta he is not even feeling cold, too 3 And there are ambiguous cases. This content downloaded from 86.177.102.88 on Mon, 18 Aug 2014 18:54:22 UTC All use subject to JSTOR Terms and Conditions NO. 2 SYNTACTIC PATTERNS IN NAVAHO AND HUICHOL 125 Constituents which function as operators Approximatem eaning 1. haah iii la ii ye6 i- don't dare 2. haa iii 16i i t?aa iv ?eiya iii-...-i ni? now let's see, how (what)? 3. -iii ki ii le? i ?at?e easily, effortlessly 4. -iii ki ii ?at?6e i ko alike, just as 5. -ii k6 i ne? behind there 6. -ii ko i sa? what if? 7. -iii ko ii ta i ?at?e possibly 8. -ii ko i t?6iya only if 9. -ii l i t44? in case I find 10. -ii la i tiini let's pretend 11. -iii le? ii ?At?66i k66 so that nothing could happen 12. la? i ii-...-lia one..., another...; some..., others... 13. -iii nti ii too iii ta i-...-iii nti ii too iii ta neither...nor... 14. -ii ta i sa? I wonder 15. -ii ta i-...-ii ta either........; for example..., or for example... 16. ta? i-...-ii is query: is it? 17. -ii tifni i nisin I mean 18. too i-...-ii ta not 19. too ii hanii i-...-ii ta nay that not 20. too i-...-iii 1 ii t44? in case I find that not 21. too v 1a iv t6? iii-...-ii ta i la it certainly is...; my but it's... 22. too iii nti i-...-ii ta not even 23. too iii t6? i-...-ii ta obviously not 24. too iii it6? ii nti i-...-ii ta obviously not even 25. too ii 6? i-...-ta is it not? 26. -i t66 iii pa?aan ii- and in addition 27. too iii 4? ii-...-i 1la I'll make sure that not 28. too ii y6 i-...-ii i it's lucky for you! 29. t?aa ii kha i- take care against 30. t?Aa ii too i-...-ii i don't! 31. t?da iv too iii-...-iv i ii ko i t?6iyg only on condition that not 32. t?aa ii too i-...-iii i ii ta just not the one (who could) 33. t?aa ii too i-...-iii i ii t44? if just not the one (who could) 34. t?aa iii too ii-...-iii k66 i ko unless 35. t?ah ii too i-...-ii ta never 36. t?ah iii too ii la? i-...-ii ta wonderw hy never 37. t?ah iv too iii-...-ii y4 i t44? before (when still not) 38. t?ah v too iv-...-iii ye ii t44 i-...-ii At?66? before...it was thus; when not...then... 39. -ii y4e i t44? when (in time past) 40. ?a4 ii-...-i nti even granted 41. ?aaan iii nee? ii-...-i nti even granted, woe is me One ambiguous case involves phrasal live there (with pA assertive mode). It is im- transforms. These function as nominals. possible, however, to posit an operator The prefix mx phrasal mode appears to be a mA i- (the one) who (...) on the basis of nominalizing operator (despite the fact that present knowledge.' it is found in independent clauses). Grimes 1 There is a parallel in Crow i (the instrument) contrasts t6wi w6na muu+y6ikaakai person which (...), apparently used for instrumental there who-used-to-live[ perhaps: he who used nominalizations: kussAzee it adheres to, i-klissaz6 to go, with the stem y6ika go, intr., kai past cork, biria door, rushzi to open, i-biria-rushtua durative, and zero third person subject] and key. Examples from Fr. John Boschi, S.J., [Gram- t6wi wana puu+y6ikaakai the person used to matical notes on the Crow language], p. [9], Ms. This content downloaded from 86.177.102.88 on Mon, 18 Aug 2014 18:54:22 UTC All use subject to JSTOR Terms and Conditions 126 INTERNATIONAL JOURNAL OF AMERICAN LINGUISTICS VOL. XXXIII A case of some complexity involves con- S junct transformations. Grimes notes that one must decide if the subject of the main clause and that of the subordinate clause are identical or different, and use a proxi- mate or obviative suffix accordingly. One must also decide if the conjunct action takes place before the action of the main clause, antecedently (and then whether the action is under way of finished [factual] or not yet begun [potential]); or if the time relation is concurrent; or if the conjunct action results from that of the main clause. The suffixes (class E4-5) can be characterized by rather formidable labels, e.g. yu obviativep otential antecedent subject, me proximate potential antecedent subject, etc., but short tags will serve as well: ta as, AA (as Joe does this... Joe does that. .), kaaku as, AB (as Joe etc., May etc.), ka after, AA, ku after, AB, me after, AA (unreal), yu after, AB (unreal), ke so that. Grimes's sample sentences, in FIG. 1 which I have detached these suffixes as en- clitics (which they may or may not be) and that. Taking ni simple future (in the main have inserted IC rankings, are: clause of 5 and 6) as an enclitic, we can 1. kuuyei + ka 2 tA 1 pAneci + ?uzei as posit -ii me i-. ..-ii ni after, AA (unreal) he (A) was walking along, A saw me, and -ii yu i-. . .-ii ni after, AB (unreal). 2. kuuyeikaa 2 +kaaku 1 pAneci + ?uzei as A was walking along, B saw me, 3. Kelley finds it unclear "why a gram- 3. nua 2 ka 1 pAn6ci + ?uzei after A ar- mar should be motivated to emphasize ele- rived, A saw me, ments that form constructions with large 4. nua 2 ku 2 pAn6ci + ?uzei after A ar- classes over those which form constructions rived, B saw me, with small classes" (loc. cit., p. 294). It 5. nua 2 me 1 pAneci + z6iyani after A might be asked why a GRAMMARIsAhNou ld arrives, A will see me, not be so motivated, as long as he does not 6. nua 2 yu 1 pAneci + zeiyani after A thereby discard or ignore more general theo- arrives, B will see me, retical considerations. It has long been 7. ne?unua 2 +ke 1 pAneci + wiitA that I recognized that a structural description has may arrive, A guides me. the property of tridimensionality. Why this Since the selection of the conjunct opera- property should not be exploited in descrip- tors is perhaps-except for the potentials-a tive and typological work is difficult to see. pragmatic affair and independent of any The tridimensionality of a structural de- verbal element, it is easy enough to posit scription can be illustrated with a diagram. -ii tA i- as, AA, -ii kaaku i- as, AB, -ii ka For the sake of illustration we will assume i- after, AA, -ii ku i- after, AB, -ii ke i- so that Huichol has these rules, admittedly imperfect: in the Archives, Crosby Library, Gonzaga Uni- CS 1 S -VP + VP2 versity, Spokane, Washington. CS 2 VP, - V1 + E. This content downloaded from 86.177.102.88 on Mon, 18 Aug 2014 18:54:22 UTC All use subject to JSTOR Terms and Conditions NO. 2 SYNTACTIC PATTERNS IN NAVAHO AND HUICHOL 127 CS3 VP2 - V2 + E2 One cannot easily say whether the re- CS 4 V, - nda he arrive, ... wards of such attention will relate more to CS 5 V2 - pAsnmecail l(+? )-zse6eiy, , . h.e .- fact-me- the theory of competence or the theory of CS 6 E, -> me after, AA (unreal) performance, but in any case I can see little CS 7 E2 - ni future justification for ignoring the typological implications that follow from the fact that These rules give us sentence 5 of section 2. The diagram directs attention to the sharp alternatives diminish from many to zero as reduction of alternatives in two places, as a sentence is generated. When nia appears this Huichol sentence is generated by the in a Huichol sentence, very little can be pre- rules. dicted about the sequel. When the me ap- Points of redundancy in grammatical struc- pears, the E1 of the diagram, it may well be that the ni or E2 is unavoidable. Those who tures have long been of interest to linguists. A major focus of attention at present is have thought of bringing game theory into the extent to which economies in describing the study of cultural systems would speak phonemes and morphemes can be effected here of 'pattern velocity'. The pattern veloc- by exploiting redundancies. In one genera- ity of some games differs markedly from that of some others. Whether and to what extent tive phonological view, for example, a seg- ment structure rule gives a constraint on pattern velocity will figure in linguistic feature composition of a segment, aiming to studies obviously is a function of how care- produce a redundancy-free segment. And a fully the various aspects of patterns in syn- tax are investigated. sequence structure rule predicts features of phonemes from features in surrounding pho- There are, then, descriptive and typologi- cal reasons for the continued study of the nemes, exploiting sequential constraints. The set of segment structure rules and sequence phrase structural components of grammars. Kelley is quite right when he insists that cer- structure rules comprises the set of mor- pheme structure rules. The dictionary matrix tain mappings of structures into structures of a morpheme is free of redundancies. It are ignored in the sort of analysis which has presents only what is idiosyncratic about been presented above. And it is quite true that immediate constituent analysis has its the morpheme, beyond what is predicted by limitations. One can analyze the constitu- general rule. The morpheme structure rules ency of The men saw the wood and still leave apply to the dictionary matrix to produce unsaid whether the sentence derives from the systematic phonemic matrix of the morpheme. The morpheme enters the set P The man saw the wood where saw can replace of rules which map representations of sen- sees, and relates to the passive The wood was tences onto the representations of the pho- seen by the men; or the sentence implies not netic level, once the morpheme structure see:saw but saw:sawed and the passive The wood was sawn by the men. This does not rules have been used. Attention to the dis- tribution of grammatical properties in this mean, however, that continued study of the and other generative phonological views patterns of phrase structural transforma- tions is bound to be fruitless. leads one to suppose that perhaps similar attention to the properties of phrase struc- view of morpheme structure rules. Still in Ms. tural descriptions would be rewarding.2 form is an important discussion and criticism of 2 See Morris Halle, The Sound Pattern of Rus- such rules, Richard Stanley's Redundancy rules sian (The Hague, 1959), pp. 19-78, for an early in phonology (Cambridge, 1966). This content downloaded from 86.177.102.88 on Mon, 18 Aug 2014 18:54:22 UTC All use subject to JSTOR Terms and Conditions

See more

The list of books you might like

Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.