LINGUIST List 31.1836

Wed Jun 03 2020

Review: Malayalam; Morphology; Semantics; Syntax: Swenson (2019)

Editor for this issue: Jeremy Coburn <>

Date: 27-Mar-2020
From: Sanford Steever <>
Subject: Malayalam Verbs
E-mail this message to a friend

Discuss this message

Book announced at

AUTHOR: Amanda Swenson
TITLE: Malayalam Verbs
SUBTITLE: Functional Structure and Morphosemantics
SERIES TITLE: Studies in Generative Grammar [SGG]
PUBLISHER: De Gruyter Mouton
YEAR: 2019

REVIEWER: Sanford B Steever,


Malayalam Verbs (henceforth, MV), a revision of the author’s dissertation, studies tense, aspect and copulas in Malayāḷam in the framework of Universal Grammar (UG). Chapter 1 introduces notions of tense and aspect used in the author’s analysis of the Malayāḷam verb. Examples from English, standing in for UG, take up the lion’s share of Chapter 1; the first Malayāḷam example appears only on page 35. Chapters 2 and 3 serve as correctives to earlier, questionable analyses of verb forms and categories in Malayāḷam. Chapters 4 and 5 undertake the analysis of common auxiliary compound verb constructions, focused on two progressive and two perfect tense forms, respectively. Chapter 6 discusses copulas.

Discussing tense from a cross-linguistic perspective, Chapter 2 argues against Jayaseelan and Amritavalli’s (2017) claims that Malayāḷam in particular, and Dravidian in general, lacks tense marking, but has aspect marking. This complements Steever’s (2018) arguments against their proposals. In her rebuttal, Swenson ascribes to JA a paradigm of Malayāḷam verbs in Table 2.2 (pp. 62, 76, 78, 133, 242) which, to the detriment of her own analysis, the author uncritically accepts. The core problem is that this paradigm asserts that the present tense form of a simple verb, e.g. ceyy-unnu ‘does, is doing’ [do-PRES], belongs to the same paradigm as, and is an optional variant of, the compound verb ceyy-unnu uṇṭǝ ‘is doing’ [do-PRES be-PRS]. Instead of treating –unnu as a present tense marker, Swenson treats it a marker of pluractionality and iterativity so that the present tense is represented by a zero morph, with ceyy-unnu segmented as /#ceyu-unnu-φ#/. However, ceyy-unnu directly and minimally contrasts with the past /cey-tu/ ‘did, was doing’ and future /ceyy-um/ ‘will do’ forms of the simple verb while the present compound verb ceyy-unnu uṇṭǝ ‘is doing’ contrasts with the past ceyy-unnu uṇṭāyirunnu ‘was doing’ and the future ceyy-unnu uṇṭāyirikkum ‘will be doing; the latter three are auxiliary compound verbs (ACVs). There is thus no need to posit a zero marker for the present tense. The misconstrual of these forms unnecessarily complicates the author’s analyses in Chapter 4.

Chapter 3 contains MV’s strongest contributions. It discusses two nonfinite verb forms, the so-called conjunctive participle (CP, Mal. vinayeccam) and the verbal noun in –atǝ (VN, Mal. kriyanāmam), traditionally and persistently (mis)represented as a gerund. Swenson insightfully characterizes CP clauses as “syntactically small,” at or below the level of vP. It will be interesting to see whether this intuition satisfies all the contexts in which the CP is used, e.g., clause-chaining, descriptive compound verbs and AVCs with such auxiliaries as varuka ‘come’ (durative), taruka ‘give to you or me’ (benefactive) and kaḷayuka ‘throw’ (completive). Swenson treats verbal nouns as a combination of an “adjectival participle” (adnominal form, Mal. pēreccam) and a number-gender marker. However, they are more transparently treated as relative clauses with a pronominal head, which may then mark number and gender.

In Chapter 4 Swenson focuses on two compound verbs she claims signal imperfective (viewpoint) aspect, var-unnu + uṇṭǝ [come-PRES be-PRES] and var-uka.y + āṇǝ [come-INF become-PRES], both roughly translated as ‘be coming’. As noted above, -unnu marks a general present tense minimally contrasting with the past tense marker –Cu/-i and the future tense marker –um. Due to mis-segementation, implying that ceyy-unnu is just a stylistic variant of the AVC cey-unnu-uṇṭǝ (see above), Swenson claims that –unnu marks [+pluractional, +iterative]. The present tense form ceyyunnu ‘does, is doing’ is consistent with such features as pluractional, iterative, habitual, generic, durative, continuous, but also appears in contexts without any of these readings, signaling simply that the time reference of the narrated event coincides with that of the speech event (where it is unmarked for the present tense to be evaluated over an interval, not at a point, of time). Specific ACVs may select a positive value for one of these features, e.g. the continuous ACV #ceyy-unnu##uṇṭǝ# ‘is doing’ [do-PRES be-PRES] is marked [+interval] and so, unlike the simple present, always has a continuous meaning. The var-uka.y + āṇǝ ACV ‘is coming’ appears to focus attention of the event/activity denoted by the main verb. The author’s claim that –uka is a progressive morpheme faces an uphill battle against the other contexts in which –uka appears but for which progressive aspect is irrelevant, e.g. the infinitive reading, the ability to take case morphology, e.g. varuka.y-āl ‘due to coming’ [come-uka-INSTRUMENTAL]. To say that –uka’s use as an imperative is “borrowed” (p. 137) is to say that the author has not yet found the form’s invariant meaning. At best, [progressive] is a circumstantial meaning of –uka: some mechanism is needed to filter out the other meanings when combined with larger structures.

Chapter 5 explores cross-linguistic variation in perfect (tense) forms, attempting to situate various Malayāḷam constructions in a taxonomy whose major values include existential perfects and universal perfects, the labels referring to quantifiers that range over temporal intervals prior to speech time. While the Malayāḷam simple verb, unlike Latin, lacks a dedicated perfect (tense) morpheme, readings of perfect tense are derived through combinations of event type, form of the main verb and choice of auxiliary in ACVs. Two varieties of the perfect are formally identified: ñān kaṇṭǝ irikkunnu ‘I have seen’ [I see-CP sit-PRES] and sārinǝ nēriṭeṅkilum anubhavam uṇṭāy(i) iṭṭǝ uṇṭō? ‘Have you ever had any direct experience?’ [sir-DAT direct=any experience be-CP put-CP be-PRES=INT]. With the first of these two forms, the perfect is often ambiguous between a progressive and a perfect meaning. In such instances, the compound ACV koṇṭ(ǝ)-irikkuka ‘be V-ing’ [hold-CP be-INF] is used to disambiguate in favor of a progressive reading. Chapters 4 and 5 assume a model that would allow us to calculate the characteristic meanings of the progressive and perfect tense ACVs from their constituent elements. The author persuasively argues that such calculations require access to morphological, syntactic, semantic and pragmatic (particularly implicatures) facets of language. This raises a risk: given the principle of compositionality, any problem in the treatment of such basic morphemes as –unnu and –uka is automatically projected to subsequent stages of these analyses.

Chapter 6 investigates copulas in Malayāḷam and compares them with other languages. It contrasts uṇṭǝ ‘be (present)’ with āṇǝ ‘be, become’. While the author claims copulas are obligatory, they are not obligatory, but normative: copula-less sentences are well attested, particularly in the spoken language. Though both copulas are components of progressive ACVs discussed in Chapter 4, Swenson misses the opportunity to test out her hypotheses by combining the meanings she ascribes to the two copulas with the –unnu and –uka verb forms in Chapter 4. It remains unclear, for example, how the analysis of uṇṭǝ as a locative, existential marker contributes to the meaning of the progressive ACV.


MV is the first major study of the Malayāḷam verbal system since McAlpin (1973) and, as such, fills in a gap in the literature on Dravidian verbal systems to advance our understanding of what factors are involved in the analysis of the Malayāḷam verb. Particularly attractive is the author’s decision to use pragmatics as well as morphosemantics to calculate the meaning of various verb forms. My remarks here include comments and course corrections, some stylistic, others more substantive. These may be useful for future projects based on the material in MV.

The text struggles with some aspects of morphology. The author wonders on p. 67 why the past is represented by a -Cu/-i morpheme: simply, both are allomorphs of the past tense morpheme. The –um in the gloss on p. 70 represents the future adnominal form, an allomorph of the adnominal marker, whose other (affirmative) allomorph is –a. In particular, the distinctions between morpheme (-), clitic (=) and word (#) boundaries are not observed. This makes it hard to tell whether Malayalam is agglutinating or polysynthetic, and leaves murky the internal structure of Malayāḷam verbs. It also obscures intermediate levels of structure, e.g. ACVs, between terminal nodes and the sentential spine that may be useful in analyzing verb forms. The hyphens in two Telugu durative forms on page 186 represent each string as a simple verb form, when both are actually compound verbs, e.g. #caduwutu# #unnaanu# ‘I am reading” [read-DUR be-PST/PRES-1S] (cf. Krishnamurti and Gwynn 1985; 170).

As noted above, Swenson aptly points out the pitfalls of accepting unanalyzed labels in her able critique of the use of ‘gerund’ for verbal noun: on p. 134 she warns against “…using terms without carefully defining them.” Her future work will wish to define such central terms as copula, mood and finiteness. Though the text talks about an “unspecified finiteness element” on p. 46, the discussion seems not to appreciate just how central this concept is to Dravidian verbal morphology and syntax; descriptions of the languages routinely categorize verbs as finite or not. Application of Steever’s (1988) characterization of finiteness to the Malayāḷam promises to yield some interesting results, viz. the past negative ceyt(u) illa ‘did not do’ [do-PST be.NEG], the present negative ceyyunn(u) illa ‘does not do’ [do-PRES be.NEG] and the durative ACV ceyyunn(u) uṇṭǝ ‘is doing’ are all serial verb formations.

In the spirit of UG, MV introduces examples from a number of other languages, e.g. Kalaallisut, St’át’mcets, Korean and Bulgarian; however, it misses the opportunity to compare and contrast Malayāḷam with its closest relative Tamil, whose verbal system has been extensively analyzed, e.g. in Steever (2005). For example, Tamil lacks copular constructions, an interesting point of comparison between the two. Though the author states on p. 59 that predicate nominals are often indicative of a tenseless language, Tamil, Kannada and Telugu all have tense as well as predicate nominals without copulas.

If each natural language instantiates UG, researchers should be able to discover facts about UG by intensive analysis of a single language, without constant reference to others. The citation of examples and analyses from other languages may offer some insights, but it also creates the opportunity for errors and garden-path effects. Fewer citations of other language examples would also free up room for the introduction of more pertinent Malayāḷam examples to help triangulate the meaning of the various verb forms. For example, the meaning of simple verb forms and the two progressive ACVs contrast with the habitual form in -āṟǝ, e.g. paṇṭǝ mutal nammuṭe nāṭṭil strīkaḷ allō tiruvātirakkaḷi kaḷikkāṟ(u) āyirunnatǝ. ‘In former times, wasn’t it women who used to dance the Tiruvatirakkali in our country? [former.times from we-GEN country-LOC woman-PL be-NEG=INT dance-HAB be-PST-VN]. Similarly, to tease out their meanings, such constructions as eŋŋane pōkān āṇǝ ‘how am I do go’ [how go-INF be-PRES] may be profitably opposed to and compared with the progressive in –uka + āṇǝ.

The attempt to conform Malayāḷam to a particular model of UG leads to some questionable claims, e.g., “Malayalam is …underlyingly basically English [p. 43].” (Perhaps, both English and Malayāḷam are underlyingly UG.) The auxiliary irikkuka ‘sit, be’ is said to appear to the “left” of the progressive (p. 187), when in actual linear order it occurs to the right. The term preposition is used multiple times in Chapter 6, but Malayāḷam has only postpositions (p. 227,ff). If linear order is irrelevant (thanks to antisymmetry), then there seems little point in labeling Malayāḷam as an SOV language and contrasting it with other word orders, as done on page 218.

MV offers no list of abbreviations used in interlinear glosses, e.g. LAM for lexical aspect modifier. It often places unanalyzed Malayāḷam morphemes or strings of question marks in interlinear glosses, e.g. –um for what is the future adnominal marker. Page 112 repeats a hackneyed claim that the adnominal form (Mal. pēreccam) –a is related to the demonstrative marker ā- ‘that’: apart from the different vowel lengths and meanings, the former is a suffix, the latter a lexical base. The References section contains a number of inadequately filled-in citations; it lacks McAlpin’s (1973) dissertation on the Malayāḷam verb.

Among the 120+ misspellings in Malayāḷam examples, the following should be noted (correct spellings are in parentheses): anweeshanam (anvēṣaṇam); tamasikka (tāmasikka); kalikk- (kaḷikk-); kuɲ- (kuɲɲ-), undu (uṇṭǝ); school-ilekku (skūḷ-ilēkkǝ); turakunna (tuṟakunna); cheyyum (ceyyum); offis-il (āphisil); asha (aṣa); jeevikkunnu (jīvikkunnu); pattiyilla (paṟṟiyilla); poyate (pōyate); vegam (vēgam); ariyum (aṟiyum); sigarettǝ (sigareṯṯǝ); veeṇappooḷ (vīṇappōḷ); naŋgaḷ (naŋŋaḷ); avaɽ (avar); malayalam (malayāḷam); veendum (vīṇṭum); maṇikkoor (maṇikkūr); raṇṭāyirati (raṇṭāyiratti); muttam (muṟṟam); pole (pōle); sankeertanam (saŋkīrtanam); vāyicch- (vāyicc-); randamooẓam (raṇṭāmūẓam); iiṭṭɘ (iṭṭɘ); acan-um (acchan=um); veliyil (veḷiyil), “aanu or undu” (āṇɘ or uṇṭɘ); kondu (koṇṭɘ ); scooter (skūṭar); caaru (kāru); nirabandham (nirbandham); aẓmkhatɘ (aẓimukhattɘ). Further misspellings in Tamil, Kurux, English and Spanish suggest a distracted copy-editor, e.g. the use of ‘&’ for ‘and’ in running text. While Malayalis can be prodigious code-mixers, the frequent use of the English words in examples rather than their common Malayāḷam counterparts is surprising: paper (prabandham), newspaper (patram, patrikai), bathroom (kuḷimuṟi, kakkhūs), zoo (mɹgaśāla), history (caritram). The lack of care over examples and their transcription may lead readers to wonder about their authenticity and question whether they may be cited. Nonetheless, at this stage of research it is easy enough to set these stylistic problems right; none seriously detracts from the issues the author has identified for study.

MV remains a work in progress to judge by some disagreements with the author’s own contemporaneous articles (pp. 87, 183) and the 15+ calls for future research at various junctures in the text (pp. 63,78, 86, 123, 129, 193, 194, 202, 207, 209-11, 222, 228, 241, 252). Nonetheless, MV sets the stage for a variety of follow-on studies in many areas of Malayāḷam verbal morphosyntax, aspect and tense. It identifies areas of morphology and syntax that may be profitably mined, and whose results will benefit Dravidianists and generalists alike. Such areas may include the dozen or more aspectual auxiliaries in the language, as well as a greater variety of ACVs. Swenson’s willingness to challenge existing analyses and to reframe them in terms of compositionality is a decided strength of MV, and she deserves full credit for identifying a series of linguistic puzzles in Malayāḷam as well as proposing the means to solve them. I look forward to continued work on this fascinating language


I wish to thank Dr. P. Sreekumar for discussing some of the Malayāḷam examples with me. Any errors in their interpretation are mine.


McAlpin, David. 1973. The Malayalam Verb Phrase in a generative matrical framework. International Journal of Dravidian Linguistics 2:252-273, 348-397.

Jayaseelan, K.A. and R Amritavalli. 2017. Dravidian syntax and Universal Grammar. Oxford and New York: Oxford University Press.

Krishnamurti, Bh. And J.P.L. Gwynn. 1985. A grammar of modern Telugu. Dehli: Oxford University Press.

Steever, Sanford. 1988. The serial verb formation in the Dravidian languages. Delhi: Motitlal Banarsidass.

Steever, Sanford. 2005. The Tamil Auxiliary Verb System. London: Routledge.

Steever, Sanford. 2018. Review of ‘Dravidian syntax and universal grammar’. LinguistList 29.3959


Sanford Steever is an independent scholar specializing in the morphology, syntax and history of the Dravidian languages. His latest book is the second edition of The Dravidian Languages (Routledge).

Page Updated: 03-Jun-2020