Review of Studies in the Compostion and Decomposition of Event Predicates
The book is a collection of nine articles connected by the common topic of event composition and decomposition. Seven articles are written from the (formal) semantics perspective, two closing chapters from the perspective of psycholinguistics. There is no explicit syntactic analysis in the book, which makes one wonder whether the idea of (de)composition of event predicates is purely semantic. However in the final chapter this idea is questioned and experimentally tested, and shown as not quite right.
The book opens with an Introduction by the editors, Boban Arsenijević, Berit Gehrke and Rafael Marín. It reviews the existing literature and presents the structure of the book, and connects previous works with the topics covered by the articles in the volume.
Chapter 2, “On the Criteria for Distinguishing Accomplishments from Activities, and Two Types of Aspectual Misfits,” is by Anita Mittwoch. It offers eight criteria for distinguishing accomplishments from activities, not all of which are equally important. At the end of the chapter we learn that the crucial criterion is telos, and when it is absent or deficient, we do not deal with accomplishments. This happens in case of DP arguments with vague indefinite quantifiers - ‘some’, ‘a few’, ‘many’, or in the predicates impossible to classify as telic or atelic. In discussing these so-called misfits, the author addresses the important issue of the imperfective paradox and mentions some of its solutions in the literature.
Chapter 3, “Lexicalized Meaning and Manner/Result Complementarity,” by Beth Levin and Malka Rappaport Hovav discusses the problem of manner/result complementarity in general and two counterexamples to this complementarity, in particular. The counterexamples are provided by the verbs ‘cut’ and ‘climb’. Both verbs are deemed polysemous; and when one sense is used, the other is suspended. It is claimed that ‘cut’ is in fact a result verb and its manner interpretation is conventional, but not lexicalized in the verb, whereas ‘climb’ is a manner verb with no component of direction of motion. The strict distinction between the meaning lexicalized by the verb and the meaning provided by its context is crucial for the analysis of verb classes.
Chapter 4, “Oriented Adverbs and Object Experiencer Psych-Verbs,” by Fabienne Martin discusses the compatibility of certain adverbial interpretations with agentive, weakly agentive and non-agentive verbs.
Chapter 5, “Two Sources of Scalarity within the Verb Phrase,” by M. Ryan Bochnak covers the issues of degree predication and degree modification. Proportional modifiers (‘half’, ‘partly’ and ‘completely’) are studied with quantity and quality scales, mostly lexicalized by the arguments of incremental theme verbs.
Chapter 6, “Interaction of Telicity and Degree Gradation in Change of State Verbs,” by Jens Fleischhauer, is another study of degree predicates. This article investigates two types of verb gradation: degree and extent. Degree modifiers also differ from each other: one class entails the truth of unmodified predication, the other does not. The main modifier of interest is ‘sehr’ (‘very’) in German with a short crosslinguistic presentation at the end.
Chapter 7, “On Adverbs of (Space and ) Time,” by Kyle Rawlins considers such adverbs as ‘quickly’ and ‘slowly’ and concludes that they only modify time, with no real connection to space.
Chapter 8, “The Processing Domain of Aspectual Information,” by Oliver Bott is based on psycholinguistic experiments conducted by the author. The conclusions of this investigation demonstrate the way the aspectual makeup of the VP is processed.
Chapter 9, “Event End-Point Primes the Undergoer Argument: Neurobiological Bases of Event Structure Processing,” by Evie Malaia, Ronnie B. Wilbur, and Christine Weber-Fox is a psycholinguistic study of argument and event structure of verbal predicates.
The book is not uniform in quality, in complexity of the analysis, or in the range of topics covered by its authors, therefore different chapters will receive different amounts of attention and criticism.
Whereas the Introduction lists nearly all the existing approaches to the central question, very few are actually raised in the volume. The more traditional discussions of event structure along the lines of Vendler's classification appear in the articles by Mittwoch and Levin and Rappaport Hovav. Both these articles are accessible to linguists of any level: they are written in clear language, and argumentation is transparent and persuasive.
Nonetheless, the article by Mittwoch contains some technical problems. The examples in (33), (34), (40) (p. 42) are unacceptable, yet not accompanied by any appropriate signs (*/#). In some places the article looks more like a handout (see, e.g., sections 2.2.7 and 2.2.8, which begin with examples and contain one and three sentences, respectively). As a result, more explanation would be necessary.
Chapter 4 first of all requires more editing, since it is full of typos and imprecision. Some French examples are not glossed or translated (pp. 72-73), on p. 77 ‘live’ is used instead of ‘leave’, on p. 78 instead of the adverb ‘cleverly’ from example (16) (p. 77) we see ‘stupidly’ throughout the discussion. In footnote 7 (p. 78) ‘itself’ refers to Geuder; on p. 84, instead of ‘TEMPORAL INDEPENDENCE’ ‘TEMPORAL INTERDEPENDENCE’ is used. When examples are taken from the Internet, the reference is very general (the Internet) (p. 88) without any concrete links. On p. 92, example numbering is interrupted by the footnote, so in the main text, (49) is followed by (51). On the same page in footnote 20, the pronoun ‘he’ refers to the adverb ‘gradually’.
More serious comments concern the interpretations of adverbials proposed by the author, and disregard of aspectual interference and the process of coercion. Thus, if one translated the French examples in (11) on p. 75 into Russian, the acceptability of subjective adverbs (like ‘cautiously’ and ‘patiently’) with what the author calls weakly agentive object experiencer psych verbs (‘attract’, ‘fascinate’, ‘find’, ‘interest’) depends on grammatical aspect: such adverbs are ungrammatical only with perfective verbs. Considering that French has the perfective-imperfective distinction as well, it would be interesting to see whether this interaction is also found. This comment also arose in connection with the example cited from Piñon (1997) (p. 90). Piñon explains the compatibility of some achievements with the progressive aspect by a meaning shift (‘he was winning the race’ = ‘he was ahead of the race’). Martin rejects this observation on the grounds that (45-a) (‘He won the race cleverly’) and (45-b) (‘He was ahead of the race cleverly’) have different meanings. But of course they do, since the author again brushes aside the aspect interference with the interpretation of verbs and (dispositional) adverbs. Martin first talks about progressive in connection to the adverb ‘patiently’ on pp. 91-92, two pages before the conclusion. This inaccuracy is also observed with respect to lexical aspect: in (52) (p. 93) ‘found out’ and ‘found’ are analyzed as the same lexical item, and the difference between the two readings is explained here by context. Thus, the detailed classification of subjective adverbs given on p. 76 and argued for in the paper runs into some confusion, which is a pity, as the paper raises very interesting and important issues.
Chapter 5 is an attempt to assign different scalar structures, testable by proportionate modifiers, to incremental theme verbs. The existence of two types of scale is demonstrated by the following ambiguous examples ((18), p. 108), in which ‘half’ can modify either the predicate (quality scale) or the nominal (quantity scale):
1. a. The meat is half cooked.
b. The glasses are half full.
The author claims that “incremental theme verbs are simple activity predicates that neither lexicalize a degree argument nor directly select for their internal argument” (p. 109). This statement requires serious argumentation. Likewise, the proposal on p. 114 that incremental theme verbs “do not show a strong attachment to their direct object” needs more elaboration. The fact that some incremental theme verbs can be used intransitively, exemplified in (23) (p.111) and (28) (p.114), does not actually mean that some object is not implied. On p. 114, two additional pieces of evidence for a weak connection between incremental theme verbs and their themes are cited from Rappaport Hovav (2008) - resultatives and prefixation with ‘out-’:
2. (29) Cinderella scrubbed her knees sore.
3. (30) Cinderella out-scrubbed her step-sisters.
In these cases, the themes are introduced by the resultative heads/prefixes and do not really look like “incremental themes” I am used to. It would be useful for the author to define what he understands under “incrementality”.
The examples in (24) (p. 111) should demonstrate that “VPs headed by incremental theme verbs do not accept the full range of degree morphology that would otherwise be expected if there was in fact an open degree argument at this level”:
4. a. ??Tim wrote the paper more than Tommy did.
b.??Tim wrote the paper too much.
In fact, what these sentences show is not that “there is evidence for a lack of an open degree argument at the VP level”, as the author claims, but rather the input made by the quantized nominal into the gradable properties of the VP. If the argument were homogeneous, like ‘poetry’, or absent altogether, no problem would arise with degree modification.
My big question concerns how the solution offered in the article could be applied to creation verbs. Either they are not incremental theme verbs, or the solution simply does not work with them. The author postulates the functional head mu (for measure), “which takes an incremental theme nominal and returns a gradable event description that is true of an event whose theme is the parts of the nominal argument, the quantity of which is equal to a degree d” (pp. 112-113). It is well known that nominal arguments of creation verbs are not yet in existence (von Stechow 2000) and cannot feed mu. Thus, applying Bochnak’s analysis to such a sentence as ‘The scarf was half knit’, we would only find a quality scale, for the nominal represents an upper closed scale and cannot be measured by ‘half’.
From the analysis on p. 113, it follows that because of the presence of mu it is impossible to say ‘eat half the apple’: one can only produce ‘half eat the apple’ or ‘eat half of the apple’. Not being a native speaker of English, I cannot be absolutely certain that the author is wrong, but I have heard the “impossible” expression more than once.
In general, the importance of mu is not immediately clear. In the diagram on p. 113, it is combined with the quantized noun (‘the apple’), on p. 115 it is claimed to be able to combine with mass and bare nouns, the difference being in the resulting scale: in the former case, it is fully closed, in the latter open. What does mu actually do, then, in addition to introducing the internal argument? (The explanations are given on pp. 113 and 115, but the question remains).
Another unclear point concerns the idea of measuring the evaluation scale provided by the verb itself. As is shown in (33) on p. 116, the verb undergoing this measuring does not have to contain the event argument (‘like’, ‘know’, ‘hear’), yet on p. 117 the complexity of certain events becomes important for their ability to be modified by evaluative ‘half’. Again, native intuitions would be helpful to tell the difference between acceptable ‘half eat’ and unacceptable ‘half melt the candle’ ((35), p. 117) from the evaluative point of view. In my non-native opinion, what influences the interpretation of each verb is world knowledge and context, not a key semantic property of a predicate. Moreover, what the author says about ‘open’ and ‘melt’ on p. 117 is applicable to some extent to ‘know’ and ‘like’ in (33) (p. 116), which are compatible with evaluative ‘half’: “there is no in-between”.
On p. 118 Bochnak uses ‘out’-prefixation as a test for the presence of the evaluative scale, and it also seems to be dependent on context. In my view, if you can create an appropriate scenario, you can use ‘out-’ with any verb, including ‘melt’, as in (37b):
5. ??Elaine out-melted Larry.
Otherwise it is equally strange to use ‘out-’ even with some ‘well-behaved’ verbs, like ‘wash’. Does ‘Larissa out-washed Elena’ sound better than (5) above ((37b), p. 118)?
Discussing the evaluative scale of ‘eat’ the author contradicts himself. The first, evaluative, reading of ‘eat’ is described as “not a prototypical eating event”, which does not track the quantity of the incremental theme argument ((4), p. 102). Further, he writes: “it is still an evaluative scale at issue in these cases, where the quality of the event is evaluated based on how much was eaten” (p. 118).
Chapter 6 differs from chapter 5 in its perspective on the nature of verbs: its proposals hinge on the fact that “the verb lexicalizes a scalar property” (p. 131). It begins with an introduction of two types of verb gradation: extent and degree. The interaction of German ‘sehr’ (‘very’) with verbs and, specifically, its influence on telicity is the subject of the article.
‘Sehr’ operates on gradable yet telic predicates, since it introduces a standard value (after the telos has been reached on the scale), which has to be reached to make the graded predication true. Thus, the definition of telicity has to be very elaborate, and the author makes an attempt to review several earlier definitions. In footnote 10 (p. 141), he mentions the notions “relative standard telos” and “absolute standard telos”, which can either be distinct or fall together. Unfortunately, little has been added to the discussion of these important notions in the main text. Table 6.1 (p. 142) presents a detailed account of event types and corresponding telos types, followed by a discussion of their compatibility with ‘sehr’.
The main problems are found in the section on cross-linguistic comparison of change of state verbs. The Russian examples in 6.7.1 (pp. 144-146) are full of typos (there are at least 8), which is largely caused by the confusion of the Latin “c” and the Cyrillic “с”. There are more serious issues too. If the situation is similar for German, this undermines some of the conclusions made in the chapter. The author studies the behaviour of such verbs as ‘increase’, ‘extend’, ‘worsen’, which he identifies as degree achievements, and ‘standardize’, ‘stabilize’ and ‘unify’, which he identifies as accomplishments. While the examples with the first group of verbs ((27), pp. 144-145) are acceptable, the examples with the second group ((28) p. 145, (29) p. 146) sound really marginal. The author found them in blogs and comment sections on the Internet, and, in my view, they represent cases of coercion similar to the change of non-gradable adjectives into gradable (‘very wooden’). I think more research is necessary on this particular class of verbs in connection with ‘sehr’-modification.
Section 6.7.2 with French examples also evokes doubt. For some reason ‘modify/change’ is assigned to the class of degree achievements (32-c, p. 148). Another moot point is the comparison of the French modifier ‘beaucoup’ to German ‘sehr’ and Russian ‘ochen’ (p. 147). Is it legitimate, considering that ‘beaucoup’, unlike its counterparts in the other two languages, can modify both the frequency and the intensity of events/states (Obenauer (1984))?
Chapter 7 discusses adverbs of space and time and develops an idea that they are basically just adverbs of time. Two crucial examples are given ((1), p. 154), where two different types of measure phrases characterize rate and temporal extent:
6. a. Alfonso ran to the park 2 miles per hour more quickly than Joanna.
b. Alfonso ran to the park 2 minutes more quickly than Joanna.
‘Quickly’ is considered to be ambiguous: in (6a), it tells something about manner, in (6b) about time. The author also says that the lexical aspect of the verbal predicate is different in the two cases above: in the first sentence it represents an activity, and in the second sentence an accomplishment.
Manner modification receives further attention. Manner adverbs represent a degree function and “the degree predication distributes over event structure” (p. 164). This complicated proposal is supported by really complicated arguments. I had the biggest difficulty with the idea of atomicity of homogeneous verbal predicates (pp. 169-170), especially when atoms of ‘running’ were presented as individual steps or even some ‘motions’ (p. 171). What would atoms of ‘working’ and other homogeneous predicates be then? The author actually addresses this question on p. 171, but he gives no answer. He needs atoms in homogeneous predicates to be distributed over by the pluractional operators, as he calls adverbs of space and time (p. 168). It is not clear why, though, since there are already multiple atoms in the structure of modified events.
Some of the analyses (e.g., (60), p. 170, (79), p. 180) remain unexplained, some questions are raised just for our information (on neo-Davidsonian analysis, p. 158), and some serious theoretical claims from previous work are mentioned as an aside (a minimal cover of the part-whole structure from Schwarzschild 1996, pp. 171-172). Instead I would prefer more explanations of the ideas the author develops, because I still could not grasp the depth of ‘winning the race’ (which he calls an “accomplishment with no internal activity component”, p. 172) representing a homogeneous join semi-lattice even if a “trivial one consisting of that event itself” (p. 172).
Another part with novel ideas is about sentence modification and narrative discourse (7.4.2, pp. 174-180). The idea of “narrative aspect” (pp. 177-178), which is of some importance for the theory developed, is never sufficiently accounted for. The author just says that it resembles perfective (p. 177), as if it is immediately clear what “perfective” means.
The author also describes events contained within narrative events (p. 178), which leaves some gaps between the events proper. For some reason, these gaps are located at the beginning of narrative events and they can be operated on by adverbs attaching high in the clause and, as a result, having inceptive readings. Thus, ‘quickly, Alfonso sneezed’ means that little time has passed after the previous event took place ((75), p. 178). So, even statives can be licensed where adverbs like ‘quickly’ appear:
7. Alfonso walked into the room. Suddenly/quickly, the students were asleep. ((77), p. 179)
I am not sure whether it correctly describes the situation in all cases. For instance, in (80) (p. 180) it looks like we could place ‘quickly’ in a different slot and the interpretation of the sentence would not change.
The chapter closes with some discussion of puzzling properties of measure phrases, where their readings more or less boil down to temporal.
Chapter 8 is one of the most clearly written articles of the volume. It takes the existing studies of aspectual interpretation and demonstrates which of them has validity in the light of four psycholinguistic experiments featuring word order variations in German. You can see what is crucial for processing aspectual information, the interaction of the verb with its arguments or the influence of adverbials.
Chapter 9 is a technical psycho-/neurolinguistics paper, which makes it difficult for laymen to understand some of the terminology - it has obviously been written for specialists within the authors’ field, therefore explanations for a number of terms (ERP, TOAL-3 LG, left anterior negativity, early left anterior negativity) must be searched for elsewhere. The article is the shortest in the collection, yet it looks neat and on the basis of elaborate analysis supports an important theoretical point: thematic roles are processed “as structural positions within an event-argument structure complex” (p. 246).
Thus, the analysis of event predicates in this volume is mostly done through the prism of adverbial modification and degree semantics, and, although there is almost no discussion of event (de)composition from syntactic positions, an important contribution is made, especially by psycholinguists in the last two articles.
Obenauer, Hans-Georg. 1984. On the identification of empty categories. Linguistic Review 4:153–202.
Piñon, Christopher. 1997. Achievements in an event semantics. In “Proceedings of semantics and linguistic theory 7”, eds. Aaron Lawson, and Eun Cho, 273-296. Ithaca: CLC Publications, Cornell University.
Rappaport Hovav, Malka. 2008. Lexicalized meaning and the internal temporal structure of events. In “Theoretic and crosslinguistic approaches to the semantics of aspect”, eds. Edit Doron, Malka Rappaport Hovav, and Ivy Sichel, 21-38. Oxford: Oxford University Press.
Schwarzschild, Roger. 1996. “Pluralities”. Dordrecht/ Boston/ London: Kluwer Academic Publishers.
von Stechow, Arnim. 2000. Temporally opaque arguments in verbs of creation. Available at http://www.sfs.uni-tuebingen.de/~astechow/Aufsaetze/Bonomi.pdf.
| ABOUT THE REVIEWER:
Eugenia Romanova holds a PhD from Tromsø University in Norway. Her thesis deals with the problems of verbal prefixation, event and argument structure and syntax-semantics interface in the Russian language. At present she is a lecturer in linguistics at a private university in Yekaterinburg, Russia.