LINGUIST List 23.4288

Sat Oct 13 2012

Review: Cognitive Science; Syntax: Takahashi (2012)

Editor for this issue: Joseph Salmons <jsalmonslinguistlist.org>



Date: 13-Oct-2012
From: William Kruger <www.krugergmail.com>
Subject: A Cognitive Linguistic Analysis of the English Imperative
E-mail this message to a friend

Discuss this message

Solid review, responded quickly to revisions and took care of things.

Announced at http://linguistlist.org/issues/23/23-1697.html

AUTHOR: Hidemitsu TakahashiTITLE: A Cognitive Linguistic Analysis of the English ImperativeSUBTITLE: With special reference to Japanese imperativesSERIES TITLE: Human Cognitive Processing Vol. 35PUBLISHER: John BenjaminsYEAR: 2012

William W. Kruger, Lincoln, NE, USA

SUMMARYChapter 1 introduces the book's overall structure: the goals and scope of thevolume, along with a discussion of the theoretical framework used -- CognitiveLinguistics and Construction Grammar -- and a summary of the data-sources used(literary sources and electronic corpora). The bulk of the chapter is dedicatedto explaining several key concepts. These include a schema-prototype model forthe English imperative (designed to account for a wide range of imperativeconstructions, both direct and indirect), along with an outline of the conceptsof "Force Exertion" and "Second Person Subject," which function to determine theprototypicality of specific examples via sets of predetermined parameters (pp.12-18). The chapter concludes with a discussion of "compatibility betweenconstructions", an element of Construction Grammar which is later used toaccount for so-called "and-conditional" and "or-conditional" imperatives (pp.19-20).

Chapter 2 centers on a frequency-based description of imperatives used inconversation, derived from a corpus of four modern works of American fiction:Sidney Sheldon's "The Sky is Falling", John Grisham's "The Pelican Brief",Daniel Steel's "Malice", and Barry Reed's "The Deception". The goal of thisquantitative study is to determine the verbs that are used most frequently inEnglish imperatives. The results are as follows: "let's" (pp. 25-33), "tell"(pp. 33-36), "let" (pp. 36-40), and "look" (pp. 40-42) show the highestfrequencies, followed by verbs such as "come/go" (pp. 43-47) and "give/take"(pp. 47-49). It is further observed that certain of these verbs frequentlyappear with a first person object in imperative constructions. Additionally, aclass of verbs (e.g. "worry", "bother") and adjectives (e.g. "silly","ridiculous") appear to systematically occur in overt negative constructions.The common feature amongst these verbs/adjectives is that they convey anadversative or undesirable outcome (pp. 49-51). Finally, various verbs and verbphrases frequently are reanalyzed as discourse-organizational markers andinterjections ("let's say/see, look, listen, come on," etc.). These observationsare part of the overall descriptive nature of the chapter, which presents adetailed and complex set of statistical figures for each verb or group of verbs.The task of accounting for the results is left until Chapter 4.

Chapter 3 returns to the key concepts discussed in the first chapter, evaluatingprevious research on the subject and describing said concepts in greater detail.Previous approaches to the characterization of English imperatives include the"anti-force" accounts of Davies (1986) and Wilson & Sperber (1988), the"illocutionary-force" accounts of Quirk, Greenbaum, Leech, & Svartik (1985) andHuddleston & Pullum (2002), and the two Cognitive Grammar accounts of Thornburg& Panther (1997) and Pérez Hernandez & Ruiz de Mendoza (2002). The authorultimately concludes that the first two are inadequate, while approaches basedon Cognitive Grammar provide some insight (pp. 67-70). The account proposedtherefore integrates elements of these models into a new approach, beginningwith the schema-prototype framework.

The Schema of the English imperative is characterized as a system of twosubevents (Event 1 and Event 2), wherein the speaker applies a certain degree offorce toward the addressee. Event 1 naturally precedes and causes Event 2, wherethe addressee conceptually performs the indicated action (pp. 73-74). Such aschema is intended to be capable of capturing all forms of imperatives,including those that do not easily fit within a "directive" category. Incontrast, the Prototype represents a more-or-less "ideal" version of theconceptual category introduced by the Schema. The prototype of the Englishimperative consists of a situation where the speaker (a Causer-Agent withinEvent 1) exerts a high degree of force toward the addressee (a Causee-Agentwithin Event 2) (p. 76). The author proposes that such a prototype be defined bytwo criteria: Force Exertion and Second Person Subject. The former is theprimary criteria, while the latter is subsidiary.

Force Exertion is calculated via a framework consisting of six parameters(desire, capability, power, cost, benefit, obligation), each of which isassigned a numerical score depending on the context of the utterance. Forexample, the parameter of "desire" can be assigned the following scores: High(+2), Low (+1), Zero, (0), Minus High (-1), or Minus Low (-2). The addition ofthe scores for each parameter yields an overall score for an imperativeutterance that characterizes the degree of force exerted toward the addressee,and this score is used to compare degrees of force across utterances (pp.76-82). The Second Person Subject criterion consists of two parameters (identityand semantic role), the values of which indicate prototypicality ornon-prototypicality. For example, the imperative "Give me a call tomorrow" isdefined as prototypical because the addressee is "individuated" (identity) andplays an explicitly causee-agentive role (semantic role). Imperatives such as"Everyone come forward" or "Shake before using", on the other hand, arenon-prototypical because the subject is non-individuated and does not play anexplicitly agentive role (pp. 87-88).

Chapter 4 recalls the results of the quantitative study performed in Chapter 2:the high frequencies of specific verbs with first person objects in imperativeconstructions, a class of adversative verbs and adjectives co-occurring withovert negation, and cases where verbs/verb phrases are reanalyzed asdiscourse-organizational markers or interjections. After reviewing these facts,the author proposes a reevaluation of the categories used to define imperativesbased upon the parameters used in the Force Exertion model from Chapter 3. Thehigh frequency of first person objects in imperatives, as well as thediscourse-organizational/interjection function of certain verbs/verb phrases, isexplained with recourse to parameters involving desire and benefit (mutual orfor the speaker) (pp. 94-99). This constitutes a departure from previousanalyses, which frequently make use of concepts of politeness, for example, indefining imperative interpretations. A similar proposal is made to account forthe co-occurrence of certain verbs and adjectives with overt negation. The factthat these elements in imperatives commonly convey an undesirable ornon-beneficial outcome is used as support for the Desire-Benefit interpretationof imperative meaning, as phrases such as "Don't bother", "Don't worry", and"Never mind" are intended to express some benefit or element of desirability forthe interactants involved (pp. 98-101). From here, the chapter turns to afurther explication of the Force Exertion parameters as a means of capturing thechoice between "direct" and "indirect" imperatives. Ultimately, elements such asurgency, politeness, or the addressee's benefit are set aside, and theparameters of Cost and Obligation are argued to constitute the primary factorsdetermining whether or not a plain imperative or an indirect directive will beused (p. 115).

Entering into a discussion of so-called "mixed imperative constructions"(imperatives involving passive, perfective, and progressive verbs), Chapter 5explicates an important element of the Construction and Cognitive Grammarapproach proposed here: the concept of "compatibility between constructions."Building upon the conception of constructions as basic grammatical units withvarying degrees of intricacy and abstractness that may be combined withdifferent amounts of success (compatibility), imperatives are characterized as acombination of constructions at the clause level: imperative, transitive, andactive. From the fact that imperative sentences using these constructions arewell-formed, it is inferred that these constructions are compatible with eachother.

Mixed-imperative constructions, in contrast, deviate from these patterns ofcompatibility. Passives may conflict with the compatibility of the "active"construction, while perfectives and progressives may conflict with theimperative construction on conceptual grounds having to do with the temporalrealization of imperatives versus perfectives/progressives. Ultimately, theconstraints on "mixed imperative constructions" are boiled down to issues ofprototypicality. In the case of passives, the conflict is conceptual, ratherthan syntactic, and it only occurs between prototypes. The passive-imperativeconstruction thus becomes more acceptable when one of the elements (passive orimperative) departs from the prototypical format. Perfectives and progressivesclash conceptually as well, since they function to locate, respectively,completed and unbounded/continuing events (i.e. before or during the time ofutterance). Perfective/progressive imperative constructions thus become moreacceptable when the reference time is shifted from its standard interpretation(i.e. from the prototype) to some future time (by the insertion of temporaladverbials, for example).

Chapter 6 discusses the phenomenon of non-directive, "conditional" imperatives:coordinate constructions consisting of an imperative and a declarative conjoinedeither by "and" (e.g. "Bring alcohol to school and you'll be suspended") or "or"(e.g. "Be careful or you'll lose your bag") (p.138). Previous approaches to theproblem of categorizing these non-directive usages are addressed using acombination of primary data (an independent survey of 10 native Englishspeakers) and secondary data (acceptability judgments gathered from previousresearch), to paint a picture of the levels of acceptability exhibited amongstthese coordinate constructions. And-conditional imperatives are determined to begenerally incompatible in embedded contexts, in distinct contrast to similar"if-conditionals". The incompatibility is tied to the overall"hearer-directedness" of the imperative construction (p. 151). Ultimately, anapproach incorporating elements of Construction and Cognitive Grammar is chosen.And-conditional imperatives are analyzed as instances of "left-subordinatingand" (Culicover and Jackendoff 1997) occurring with a prototypical ornon-prototypical imperative form that, crucially, cannot express maximallystrong force (p. 158). Or-conditional imperatives, conversely, are analyzed asinstances of "asymmetric or" (Lakoff 1971) combined with a subclass ofprototypical imperatives which must, crucially, exert maximally strong force (p.163). Throughout the analysis, "compatibility between constructions" resurfacesas a means of characterizing the observable incompatibilities amongst thesestructures (pp. 157-158, 159, 163). The Force Exertion model is also applied tothe observable differences in the allowable degree of force, yielding the resultthat imperatives in and-conditionals cannot appear with a maximum force, while(in what is characterized as a form of "complementary distribution", p. 167),imperatives subordinated with asymmetric-or are unrestricted in this respect.

Chapter 7 turns to imperatives appearing in subordinate, "concessive" clauses;that is, clauses introduced with adverbial conjuncts such as "(al)though, except(that)" or "even though". Some initial observations are made in reaction toLakoff (1984), centering around the distribution and possible interpretation ofimperatives when they are embedded in these types of clauses. Clauses introducedby "(al)though" and "except (that)" allow directive interpretations, while thoseembedded with "even though" do not. In addition, imperatives in reason clauses(introduced by "because/for") require a rhetorical assertive interpretation.These observations are established by recourse to a range of previous research,as well as data gathered from the British National Corpus. To account for thesefacts, the concept of "compatibility between constructions" is again brought tobear, examining the integration of imperative and subordinate clauseconstructions. The compatibility of directive-force imperatives embedded with"(al)though" and "except (that)" is explained as a result of a lower degree ofintegration between the imperative and the main clause (pp. 186-188), while theincompatibility of such clauses with "even though" is ascribed to a conceptualconflict between the non-rectifying concession of "even though" and the standardrectifying concession of imperatives (pp. 191-194).

Chapter 8 expands the theoretical approach of the previous chapters to anexamination of Japanese imperatives. The chapter begins with an overview of theprinciples of Japanese imperative verb forms, specifically three forms forcommands (siro, sinisai, sitemiro) and two forms for requests (sitekure,sitekudasai). The goal of the chapter is to determine the extent to which aForce Exertion model can be applied to imperatives in Japanese and also to gaugethe applicability of the "compatibility between constructions" approach to thespecific case of passive imperatives in Japanese. The author comparestranslations of Japanese and English imperatives, and ultimately concludes thatimperatives in Japanese express the full possible range of variation within theFE scale (pp. 201-08). In the case of passive imperatives, constructions withthe command form sitemiro ("try") are found to be much more broadly acceptablethan other forms, and this is attributed to an inherent aspect of sitemiro thatallows it to deviate drastically from the prototypical imperative pattern (pp.214-216).

Chapter 9 concludes with a recapitulation of the content of the previouschapters, enumerating the various proposals and conclusions made, followed by adiscussion of the implications of the unified approach presented forimperatives, in the interest of further expansion and research. The chaptercloses with a statement concerning the need for further cross-linguisticresearch into the application of the Force Exertion/Second Person Subjectframework in order to further determine its overall validity.

EVALUATIONIn this work, a fully-described framework for classifying imperatives ispresented, well within the bounds of Cognitive Linguistics and ConstructionGrammar. The author's detailed statistical analysis of data certainly lends tothe book's overall value. The presentation and theoretical discussion of saiddata is, however, very dense and technical at points (Chapters 2, 4, and 7 inparticular), and readers must therefore be familiar with a range of advancedtheoretical concepts. In contrast, the proposal of the Force Exertion frameworkin Chapter 3 is very clearly and concisely described, and the outline providedin Chapter 1 is immensely valuable as a reference for keeping track of theprogress of the book. Overall, the framework of Force Exertion/Second PersonSubject and the concept of "compatibility between constructions" are bothconsistently applied to the range of phenomena examined.

Any further criticism must therefore come from a theoretical direction, e.g.involving the cross-linguistic value of the system proposed. It cannot be deniedthat a scalar system such as that proposed in the FE model is certainlydescriptively adequate. However, in a book that presents an otherwise extensivedescription of imperative structures, there seems to be a lack of discussionconcerning the formal, internal nature of imperatives. That is, what (ifanything) actually distinguishes an imperative construction from any othercomparable construction? Such a question may, however, prove unimportant withina Construction Grammar framework, which envisions constructions as arbitraryelements in a given language. If constructions, as units of grammar, are assumedto be arbitrary, the formal difference between an imperative construction and,say, a declarative construction is trivial, and the appearance of either ofthese constructions in any particular language is also (ultimately) arbitrary.The focus, in this work, on the interpretation and statistical patternsexhibited by imperatives in English is therefore consistent with the assumptionsof Cognitive Linguistics. By that measure, the framework of analysis proposedhere is certainly a valid computational means of describing the force ofindividual imperative utterances and provides a more fully realized account ofa, historically, very slippery issue: the classification of the possiblerealizations and interpretations of imperative structures.

A proofreading evaluation yields nothing beyond a few minor errors, as follows:"coexitst", p. 19, the second sentence of the second full paragraph; "To taken",p. 27, the second sentence after the list of examples in (3); "As a matter ofact", p. 31, the second sentence after the list of examples in (9) on theprevious page; "operation" for "operations", p.134, the last sentence of thelast paragraph; "imperative' " for "imperative's", p. 145, the first sentence ofthe second paragraph in section 6.3.

The framework proposed in this work is a valuable addition to the field ofCognitive Linguistics. It provides insight into the elements by which imperativeutterances can be classified, along with tools for comparing these utterancescross-linguistically. Ultimately, such a framework can serve as an importantbasis for further research in the area of imperative constructions.

REFERENCESCulicover, P. W. & R. Jackendoff. 1997. Semantic subordination despite syntacticcoordination. Linguistic Inquiry 28(2). 195-217.

Davies, E. 1986. The English Imperative. London: Croom Helm.

Huddleston, R. & G. K. Pullum (eds.). 2002. The Cambridge grammar of the Englishlanguage. Cambridge: Cambridge University Press.

Lakoff, G. 1984. Performative subordinate clause. Berkeley Linguistics Society10. 472-480.

Lakoff, R. T. 1971. If's, and's and but's about conjunction. In C. J. Fillmoreand D. T. Langendoen, eds., Studies in Linguistic Semantics: 115-149. New York:Holt, Rinehart and Winston.

Pérez Hernandez, L. & , F. J. Ruiz de Mendoza. 2002. Grounding, semanticmotivation, and conceptual interaction in indirect directive speech acts.Journal of Pragmatics 34. 259-284.

Quirk, R., S. Greenbaum, G. Leech, & J. Svartik. 1985. A comprehensive grammarof the English language. London: Longman.

Thornburg, L. & K-U. Panther. 1997. Speech Act Metonymies. In W.-A. Lierbert, G.Redeker, and L Waugh, eds., Discourse and perspective in Cognitive Linguistics:205-219. Amsterdam: John Benjamins.

Wilson, D. & D. Sperber. 1992. On verbal irony. Lingua 87. 53-76.

ABOUT THE REVIEWERWilliam Kruger is a freelance linguist and editor. He received his M.A. inLinguistics from Arizona State University. His research interests includehistorical Germanic linguistics, theoretical syntax and phonology.

Page Updated: 13-Oct-2012