论文部分内容阅读
摘 要:认知语言学语法原型理论框架下,及物性作为一种语法范畴,是具有典型性的,动词及物性被认为是一个语法原型。本研究对典型及物性观点进行反思和批判,并证明这一观点所引发的结论的不全面性。原型范畴仅在语用解释层面起作用,只是对现象的菱镜似的解读,解释不了及物性问题。
关键词:及物性;认知观;反思
1 Introduction
Cognitive linguists consider that language category has its prototypical organization structure. The prototype is regarded as the optimal paradigm of a concept in a general way. The basic meaning of the grammatical prototype rests with certain grammatical marker, construction, and event type that connect with some morphological features in different languages. The prototypical transitivity event has only two elements: one participant is concerned with some action volitionally. If the transitivity is treated as a continuum, the prototypical transitivity and the intransitivity events constitute the two ends of the continuum. It is stated that the transitivity is a semantic concern instead of syntax aspect, if it is looked at from a higher, abstract level; transitivity is also a grammatical prototype.
2 Exploration of Langacker’s View of Transitivity
2.1 Langacker’s View of Transitivity
Langaker (1991, p. 285) proposes the model of the ‘the normal observation of a prototypical action’ in discussing transitivity, sketched in Figure 1, based on the following observations:
A viewer (V) observes an event from a vantage point external to its setting. The event consists of two participants of an agent and a patient represented by the two circles respectively. The transmission of energy is depicted by a double arrow whose direction clearly shows the energy transferred from agent to patient. The squiggly arrow indicates the internal change of the patient as a result of the energy transmission. When people are in the process of coding the event by using linguistic expressions, they show the corresponding relationship between the cognitive model and the structures of clauses. For instance, the speaker and addressee correspond to the viewer, the verbs to the energetic interaction, other grammatical elements to participants or setting. Langacker claims that the canonical event model reflects the prototypical value of a finite transitive clause in that it conceptually correspond to the archetype of an energetic interaction and structurally manifests the fullest array of properties that are specifically clausal. As a notion for the characterization of the finite clause prototypes, transitivity is seen not in terms of nominal occurring in a particular structure configuration, rather a matter of degree. In line with Hopper and Thompson (1980)’s work on the proprieties with reference to transitivity, Langacker believes that all those conceptual factors can be identified as facets of the canonical event model which tie them together and provide a basis for the prototypical notion of transitivity. For example, Jim broke my cup. The organization of this sentence reflects the canonical event model and it is quite high in transitivity because it corresponds to the prototype regarding multiple factors of the model. The nominal complements of a relational predication are often referred to as its arguments. The arguments of a prototypical transitive clause represent distinct, clearly delimited participants that are sharply differentiated from each other, from the profiled event, and from the countless other entities that could conceivably participate in their stead. Found in all languages, however, are grammatical devices that accommodate departures from this canon. Such devices permit the efficient coding of events in which the participants are not in fact distinct, or where the speaker is unable to specify an argument with any precision. The resulting expressions usually have diminished transitivity, even to the point of being fully intransitive (Langacker, 1991, p. 362).
The conception of two participants thus interacting is a central feature of the canonical event model, the prototype for transitive clauses. The participants of a prototypical clause are therefore separate and discrete physical objects that exist both prior to the profiled event and independently of its occurrence.
Our immediate interest lies with the broad class of non-conforming expressions in which a participant is not wholly independent of the designated process (Langacker, 1991, p. 362). A minimal lack of independence is exhibited by participants that have no prior existence but are rather created by the event itself: build a cabin; knit a sweater; carve a figure; paint a masterpiece; draw a diagram; light a fire; make dessert; cause a traffic jam; devise a plan; write a letter; compose a sonnet; strike a deal; create a consensus; invent a new technique; etc. Hopper refers to such participants as “effected” objects and surveys their properties in English and other languages. He finds that effected objects are seldom distinguished from “affected” objects by any special marking; they may however show behavior symptomatic of lesser transitivity. Observe, for instance, that He cut a slice of cake permits an effected-object interpretation whereby the cutting brings the slice into existence. However, adding the particle up imposes the affected-object construal: He cut up a slice of cake can only mean that a pre-existing slice was partitioned. Hopper notes that cut up represents a higher degree of transitivity than cut alone, for it generally indicates that the action affects the object in its entirety (e.g. He cut up the meat vs. He cut the meat).
2.2 Analysis of Langacker’s Argument on the Degree and the Meaning of Transitive Clauses in Transitivity A clause that has both a subject and a direct object is said to be transitive. In recent years, the traditional notion of transitivity has been investigated in considerable depth and has proved to be both subtle and complex (Hopper and Thompson 1980). There are cases where a verb is arguably transitive despite the lack of an overt object. For instance please in Roger always aims to please and dried in Peter washed the dishes and Marsha dried.
More seriously, the occurrence of a nominal in direct-object position does not invariably render a clause transitive, as determined by such traits (depending on the language) as case marking, agreement, passivizability, and special verb morphology characteristic of transitives. For example, although the sentences in (1) have post-verbal nominals that appear to be direct objects (on the basis of position and the absence of a preposition), the infelicity of the corresponding passives in (2) calls their transitivity into question.
(1) (a) After his divorce, Jason became a new person.
(b) That little boy resembles my nephew.
(c) Tuesday witnessed yet another gang-related slaying.
(d) The reactor’s cooling system burst a pipe.
(e) The former emperor died a peaceful death.
(2) (a) *After his divorce, a new person was become by Jason.
(b) *My nephew is resembled by that little boy.
(c) *Yet another gang-related slaying was witnessed by Tuesday.
(d) *A pipe was burst by the reactor’s cooling system.
(e) *A peaceful death was died by the former emperor.
Such considerations have made it increasingly apparent that transitivity is not definable just in terms of nominals occurring in a particular structural configuration. It is instead a matter of degree and depends on the meaning of the clause as a whole (Langacker, 1991, pp. 301-302).
Works such as Hopper and Thompson, Hopper, and Rice have shown that a substantial number of conceptual factors contribute to transitivity. Among the properties characteristic of a prototypical transitive clause are the following: (1) it has two participants expressed by overt nominals that function as subject and object; (2) it describes an event (as opposed to a static situation); (3) the event is energetic, relatively brief, and has a well-defined endpoint; (4) the subject and object represent discrete, highly individuated physical entities; (5) these entities already exist when the event occurs (i.e. they are not products of the event); (6) the subject and object are fully distinct and participate in a strongly asymmetrical relationship; (7) the subject’s participation is volitional, while that of the object is non-volitional; (8) the subject is the source of the energy, and the object is its target; (9) the object is totally affected by the action. This is not a random list. As pointed out by Rice, all these factors can be identified as facets of the canonical event model it is this model that ties them together and provides a coherent basis for the prototypical notion of transitivity. These sentences in (1) are quite low in transitivity because they deviate from the prototype in regard to multiple factors (Langacker, 1991, p. 302). While, In She likes it, for example, the profiled mental interaction (of the form EXPER→ ZERO) is construed as an abstract analog of the energetic interaction (AG→ PAT) that is prototypical for transitivity, so it is coded linguistically as a transitive clause (Langacker, 1999, p. 31). This we cannot afford to agree to. According to the above parlance, the sentences in (1) have post-verbal nominals that appear to be direct objects; the infelicity of the corresponding passives in (2) calls their transitivity into question, which is really arbitrary. Firstly, (1a) actually differentiates from the transitive verb resembles in (1b) since became itself has two lexemes; one is as transitive verb while the other is taken as the intransitive in traditional grammar. And a new person should be regarded as the intransitive whole, for a new person is a part of meaning of the intransitive and the complement of the complete meaning of the intransitive became. As for the intransitive verb, became indicates ‘come to be’ and as for the transitive one, it implies to ‘be appropriate or suitable to’, which shows the complex of the verb became. Therefore, we have to admit that the passive expression of a new person was become by Jason is eventually not reasonable semantically and is to be deleted, which is not contradictory to the infelicity of the corresponding passive in the transitivity problem without question. It is obvious that those listed examples are not representative in fact to reason the problem. Secondly, whether or not the passive pattern is overt or explicit poses no challenge to the transitivity. As a matter of fact the transitivity problem is the abstract category involving the aspects of language performances undoubtedly, but the transitivity is hierarchical and also recursive as the syntax deals well with this matter. And the absence of the passive form is not equal to the non-existence of covert structural computation. When it is manipulated by our mental operation, the appearance and dormancy can achieve a balance through struggling compromise or tuning up, that is to say, the result is the representation of the passive form or the vacancy, which is possible as well. There is a tension between the prominence and recessiveness, which reaches the higher unification by dint of the law of thought. If the practical phenomena take the place of the rank of reason, it brings out the chaos of the realization. The infelicity of the corresponding passives in (2) calls their transitivity into question finally has its view of the important overshadowed by the trivial, as we can not take the place of abstractness by means of the concreteness. The third, the infelicities of the corresponding passives in (2), all phenomena are merely the language representations since absence is another way of expression. And the correspondence between form and function or meaning is not the simple one-to-one relation, as our expressions are the shadow of the shadow of the idea or logos, in other words, our language performance is the replica of the language mechanism operation. Looking back at (1a), the verb became itself containing the transitive and the intransitive usages are the inspissations of the transitivity problem, and it shares the functions of the two poles by virtue of one form, that is, it is the isomorphic exhibit. We claim that the used or accepted passive transformation, or otherwise can not cancel the nature of the verb, that is to say, the transitivity problem is the category domain and the yes-no question, but irrelevant to the degree or the deviation from the prototype in regard to multiple factors. As the multiple factors themselves are impressed by the experiential and phenomenal operation and too much subjective, undoubtedly, they are not of the qualification of scientific rigidity and easy to collapse by means of few counterexamples. The details of falsification is similar to the latter, here we leap over. As to the voice of high or low in transitivity is ridiculous in practice, for the sensual, intuitional and psychological judgment or the data from the statistic operation is not persuasion whenever and wherever. After all, it is fluid or in flux and it is difficult to cater for all tastes, which is not rational, not persuasive and not convincing. Langacker has neglected other possible interpretations of an expression. We combine words into sentences with pre-determined lexical meanings of each word, and know which meaning is intended. If it demands, the null or empty elements could be complemented through our construal process and understanding. He takes radial instead of hierarchical relation. Further, there is no ‘surplus’ usage accruing to a verb no matter transitive verb or intransitive one that is not attributable to its principle and parameters mechanism. Universals may be breached but they need not be present. It is stated that the rules of language do not consider simple linear order but are structure-dependent. Structure-dependency is a discovery about the nature of human language in general, and a principle of Universal Grammar.
What distinguishes transitive verbs from intransitive ones is the syntactical structure instead of semantic values. And passivity cannot be used as the only standard of the transitives. Since passivization is a manipulation and one sort of representational expressions in transitivity, which makes active sentences such as Tom killed him into passive sentence He was killed by Tom. In this process, the object of the active sentence is turned into the subject of the passive sentence. The two sentences are different in their respective viewpoints. As to the above two sentences, in the active aspect, its viewpoint is neutral, on the other hand, in the passive sentence the speaker describes the event only from the patient’s view. However, not all the events or states expressed in transitive sentences can be described from the patient’s point of view. This is the matter of perspective or construe in humans’ eyes.
In this case, it is construed and checked as the erasing or absence of passive usage. However, the desuetude in this situation does not mean the non-existence of the concealable language mechanism. It is pivotal that grammar is there. It works on its own and with resources from relexification, which has reified through interaction between language faculty and language use (Zhao Yanchun, 2006, p. 19). In communication, we often omit many elements of the utterances as default values like the given example dried in Peter washed the dishes and Marsha dried which has been mentioned at the very beginning of this chapter, but have no reason to veto the transitive nature of the verb dried, as its indicated object is covert or not necessary to appear in virtue of the relevant and economic principle of expressions. Again rethink the mentioned examples Jerry bumped into John and John was bumped into by Jerry, which are mentioned in chapter two already. Through the further syntactic analysis, we regard that was bumped into itself is a verb phrase and need not constituent-select or category-select the object as a norminalized phrase, and it is the logical objected. And it is the verb phrase was bumped into that semantic-selects the theme John, not the patient. Therefore, through the reference of that deep syntax analyzing, we point out that syntax has no semantic meaning or no semantics concern and syntax is autonomous, which can elaborate the passivity matter and give a well done.
And in The Norman defeated the Saxons, the transitive verb defeated category-selects an NP the Saxons, while semantic-selects two θ roles, one is the agent The Norman and the other is the patient the Saxons. In the same way, in The Saxons were defeated, were defeated as one node; that is to say, one verb phrase has no category-selection and semantic-selects one θ role The Saxons as the theme. Through the syntactic computation, the passivization phenomena can be made clear. It is stated that phrases are members of syntactic configuration, and logical form, partial representation of meaning is determined by grammatical structure. In syntax, phrase grammar is rather than word grammar as were defeated is taken as a constituent in syntactic operation. The links between syntax and the lexicon should be definite and important because structure-dependency also affects the interpretation of sentences. Lexical structure must be represented categorically at every syntactic level. The transitivity problem can be explicit by the way of syntactic modular operation of language mechanism.
From another perspective, meaning might be the basis and warrant of understanding classification to some extent, but not the essential or the fundamental level compared with the grammatical structure’s criterion. Dividing into transitive verbs and intransitive verbs serves syntax analysis, so the standard of classification should be the formal one, that is, objects can follow the transitive verbs directly while the intransitive verbs cannot in the semantic way. The coming question is what the objects are, that is to say, what elements can be used as the criterion of transitivity’s ranking. It is worthy to use for the reference differentiating the complement or not in the formal valence grammar. When deleting some elements, whether the left sentence structure accords with grammar or not, if it is answering for it, the erasing components may have its complement or free expository points; if not, the erasing is to be followed by the complement. In so many words, the formal valence view for distinguishing transitive and intransitive verbs seems an undoubtedly newly approach. It is clear that semantics must be distinguished from syntax with no strings attached, since they are two different modules governing their own operational ranges. On occasion, semantic meaning may be neglected when it is not too necessary, and it is covert or implicit. As syntactic structure is not indispensably determined by semantic meaning but able to make clear the semantic value, we can claim that syntax is self-independent and autonomous. Such as love, the expressed same semantic meaning might be a noun or a verb as well in syntax, which discloses its kind that is determined by the syntactic structure, irrelevant to the semantic aspect. The following (3) can present a clear explanation.
(3) a. I love him.
b. My love for him is a mistake.
c. Love is great.
In (3), love in all those three sentences shares the same semantic meaning undoubtedly, while the distinct class features that (3b) and (3c) compared to (3a) possess, the former two belong to the noun category and the latter the verb in the totally different syntactic structures from the syntax perspective. Since in syntax the structure-dependency determines the various syntactic features of love, irrelevant to the meaning, it is stated that semantics and syntax are different modules and syntax is autonomous without question.
Returning to Langacker’s evidences, we have our rational thought. As the case stands, the infelicity of the corresponding passives in (2), that is to say the forms of (2) make some information default, which is not the comprehensive problem or the semantic understanding to the quick, but the matter of non-conventionalization or the result of non-phylogeny, that is, they are not taken in for the moment, which does not mean they bear no potential competence in syntax.
Furthermore there is a tension between transitives and intransitives, and the minimal pair of them exists. Any expression is the representation of the structure-independency in language; therefore, all the absent information of (2) shows the recessive application of transitives in (1).
As to the displayed hypotheses, are all those nine factors really all-sided to be as principles or the mode of checks and balances in transitivity? However, it is the fact that it will have our view of the importance overshadowed by the trivial in conformity to their asseveration unfortunately. Firstly, it is based on the experiential representation of language, not satisfying stringency. Therefore, meaning is involved to analyze and induct the transitivity, which is inadequate to the operation of transitivity. Since semantics and syntax are independent of each other, the former is the expressional system of language meaning and the latter is the computational system of language mechanism, and there is no absolute and necessary correspondence relation. The second, its depth of field is rather confined by limited expressional resources, if it sounds with reason, which is considered as a local availability, because in fact there exist a large number of disproval exemplifications to refute those statements. It points out those properties characteristic of a prototypical transitive clause collapse in the face of indelible convincing counterexamples. Thirdly, similar to the above nine factors, Taylor also fetches a series of similar hypotheses. We adopt the logic tool to challenge Langacker predication by falsification and also Taylor’s collapses of itself finally. 3 Discussion of Taylor’s Assertion of Typical Instantiations
3.1 Taylor’s Arguments
Taylor (2001, p. 206) brings forward the prototypical hypothesis in transitivity. The followings are typical instantiations of the transitive sentence construction.
(4) The child kicked the ball.
(5) John moved the table.
(6) Mary killed the intruder.
The syntactic properties of the construction may be represented by the formula NP1VTRANSNP2, where NP1 and NP2 stand for the subject and direct object, and VTRANS is a transitive verb. In its prototypical instantiations, both NPS have specific reference, while the verb is realis, i.e. affirmative and indicative, and in a reporting tense (either present or past). Drawing on Lakoff (1977) and Hopper and Thompson (1980), we can list at least eleven semantic properties of the construction, in its prototypical instantiations.
(a) The construction describes events involving two, and only two participants, encoded by the subject and direct object NPS respectively.
(b) The two participants are highly individuated, i.e. They are discrete, specific entities (from this it follows that both the NPs in the construction have specific reference), distinct both from each other, and from the background environment.
(c) The event is initiated by the referent for the subject NP, i.e. by the agent. Responsibility for the event thus lies exclusively with the agent. Furthermore, the subject NP is the sentence topic; the subject is what the sentence is about.
(d) The agent acts consciously and volitionally, and thus controls the event. Since consciousness and volition are typical human attributes, it follows that the agent is typically a human being.
(e) As a consequence of the agent’s action, something happens to the patient, i.e. the referent of the object NP. The effect on the patient is intended by the agent. Often, though by no means necessarily, the patient is inanimate.
(f) After the occurrence of the event, the patient is in a different state from before the event. Usually, the difference is one which would highly perceptible to an overlooking observer.
(g) The event is construed as punctual. Even though the event necessarily has temporal extension, the internal structure of the event, and the intermediate states between its inception and termination, are not in focus.
(h) The agent’s action on the patient usually involves direct physical contact, and the effect on the patient is immediate. (i) The event has a causative component —the agent’s action causes the patient to undergo a change.
(j) Typically, agent and patient are not only clearly differentiated entities; often they also stand in an adversative relationship.
(k) Finally, the events reported by the construction are real, not imaginary, hypothetical, or counterfactual. Hence, central instantiations of the construction are realis.
(Taylor, 2001, p. 207)
3.2 Analysis of Taylor’s Arguments
Let us see whether Taylor’s view is tenable. Now we shall analyze Taylor’s arguments length by length. Let us use Taylor’s standpoints for our counter-argument.
Taylor regards that prototypical instantiations of the described events involve two, and only two participants, which is not reasonable actually according to plentiful basic language facts, like She gave me a book. In this case, obviously there is not merely a simple question of the subject and the direct object. Nevertheless, we know quite well that one direct object and one indirect object follow the transitive verb gave. Therefore, (a) is not taken as the respective criterion of the prototypical transitivity.
As for (b), again Taylor has overlooked the language fact, such as You should help each other and I hate myself which exists everywhere. Considering that fact of language, nobody I suppose can agree with Taylor, for you and each other or I and myself are the same reference or entity, that is, the latter each other or myself refers to the former you or I.
As regards statement (c), which asserts that the agent is responsible for the whole event, we agree that the process of the event absolutely connects with the subject, however, he says the subject NP is the sentence topic is not persuasive for every aspect.
Explicitly, Taylor’s fallacy is that he has over-generated the entry to transitivity following (c). For instance, the expression Happiness gave me a lot of inspiration is a reasonable and acceptable transitive clause. However, inanimate Happiness is taken as the subject, i.e. the agent, which is rather legal after all.
Taylor’s argument is so subjective that it is full of contradictions. Again look back at the former example Happiness gave me a lot of inspiration. With regard of this case, me is one of the patients logically and an indirect object and a human being without question, which is rightly in contradiction with the assertion that the patient is inanimate. With regard of this case, Taylor’s argument seems explicit; however, he mixes the different levels of language. The operational mechanism and the understanding realizations are not the same process. Again like I hate myself, the internal change is not grasped by means of the representation of language, and the interpretations of the understanding process are various.
Taylor’s argument is too arbitrary and unilateral in (h). Firstly, the sayings ‘direct physical contact’ is rather limited and insufficient as the agent’s action, which deals with the psychological and mental recognition, considering the effect is quite changeable according to different people and different situations. Secondly, The effect on the patient is immediate is the only tip of real facts. Look at the following example I see what you mean, there is no direct physical contact and the effect is involved in the patient, not as simple as the immediate reaction after all.
As for (i), what kind of mistake Taylor makes is that he comes to conclude through the limited experience. How to confirm that the agent’s action causes the patient to undergo a change is reasonable, which may be really hard. For instance, in I miss you very much, what change is implied, it is not easy to answer.
In the same way, as to (j), we can find at least one exception to falsify Taylor’s arguments. Since agent and patient are not clearly differentiated entities, their relationship is not sure to be only the case of adversativeness. In fact, the relationship is not simple just as what he says, but is plentiful according to colorful language resources. Such as in I enjoy myself, myself is the anaphor of I, so they are tied by the same relationship.
In (k), Taylor takes his every effort to maintain his opinion of prototypical transitivity, however, he overlooks the mechanism of language’s syntactic operation and the law of human’s logic thought. He overgeneralizes and draws a conclusion of that so-called central instantiations.
More fundamentally, Taylor’s problem as well as that of other cognitivists lies in the negligence of the nature and/or mechanism of the human speech, which results in the inaccuracy of taking the performances or representations of language for granted as the maxims of the transitivity entry. Taylor’s claim of transitivity has theoretic weakness resulting from their abandoning of the classical category theory. Now that the above detailed rules cannot be verified, but to be falsified. 4 A Non-Prototypical Transitivity Verdict
In short, the above linguistic evidence stringing along with their positions pose no threat to the classical account of transitivity. If applications are taken as the transitivity category itself, our lexicon would not have such a capacity, and communication would be impossible. This problem has been elaborated by Aristotle. Evidently, the above hypotheses are not necessary.
In light of their ‘It goes without saying’, the prototypical transitive clause possesses most of the semantic properties, thus most representative of transitivity category. It is evident that all of these semantic features have their experiential basis, corresponding to Langacker’s canonical event model. What is more, they believe that this fact also coincides with the interactional nature of properties that are seen to determine the prototype of a category, and such bodily interaction with the outside world Lakoff believes from clusters in our experience, and prototype can reflect such clustering. However, how to judge the highest level of transitivity in terms of those rules?
We have to point out their declarations resting on physical, psychological, and mental refraction, which cannot hit the core of problem. Although, relying on realistic embodiment, we really construe the world and achieve our realizations to some extent, which merely acts in the superficial stage whatever happens. Since the world is not what we have seen, language is representational, and category is abstract; this eventually educates and leads us to go deep into the nature of phenomenon instead of concrete and experiential things. Thus problems are coming to us. Is the determination of prototype definite and clear? Are transitives prototypical and the pre-appearing to be undoubtedly prototypical as well? Then which one is the pre-existence, transitives or intransitives? Is it possible that as long as it is more applying frequency, then more prototypical? Frankly speaking, prototype category is not able to answer these fundamental questions of transitivity problem, and then the prototypical transitivity is on the verge of imminent death. Therefore, the so-called prototypical transitivity wherewith to shake the ground of the foundation of the classical category is but a whim. Considering the analysis of Langacker’s arguments and Taylor’s prototypical hypotheses in transitivity cannot hold, the non-prototypical transitivity should be affirmed ultimately.
[References]
[1]Hopper, P. & Thompson, S.A. Transitivity in grammar and discourse. Language. (1980), 56.
[2]Lakoff,G. Linguistic gestalts. In CLS (13), Papers from the 13th Regional Meetings. Chicago: Chicago Linguistic Society. 1977.
[3]Langacker, R. W. Foundations of cognitive grammar (vol. 2): descriptive application.. Stanford: Stanford University Press. 1991.
[4]Taylor, J. R. Linguistic categorization: prototypes in linguistic theory. Beijing: Foreign Language Teaching and Research Press. 2001.
[5]赵彦春.语言哲学(课程).2006.
本文系2012年度中央高校基本科研业务费专项资金资助(supported by the Fundamental Research Funds for the Central Universities)“及物性原型观的批判性研究”(2012RW016)
(作者单位:华中农业大学外国语学院, 湖北 武汉 430070)
关键词:及物性;认知观;反思
1 Introduction
Cognitive linguists consider that language category has its prototypical organization structure. The prototype is regarded as the optimal paradigm of a concept in a general way. The basic meaning of the grammatical prototype rests with certain grammatical marker, construction, and event type that connect with some morphological features in different languages. The prototypical transitivity event has only two elements: one participant is concerned with some action volitionally. If the transitivity is treated as a continuum, the prototypical transitivity and the intransitivity events constitute the two ends of the continuum. It is stated that the transitivity is a semantic concern instead of syntax aspect, if it is looked at from a higher, abstract level; transitivity is also a grammatical prototype.
2 Exploration of Langacker’s View of Transitivity
2.1 Langacker’s View of Transitivity
Langaker (1991, p. 285) proposes the model of the ‘the normal observation of a prototypical action’ in discussing transitivity, sketched in Figure 1, based on the following observations:
A viewer (V) observes an event from a vantage point external to its setting. The event consists of two participants of an agent and a patient represented by the two circles respectively. The transmission of energy is depicted by a double arrow whose direction clearly shows the energy transferred from agent to patient. The squiggly arrow indicates the internal change of the patient as a result of the energy transmission. When people are in the process of coding the event by using linguistic expressions, they show the corresponding relationship between the cognitive model and the structures of clauses. For instance, the speaker and addressee correspond to the viewer, the verbs to the energetic interaction, other grammatical elements to participants or setting. Langacker claims that the canonical event model reflects the prototypical value of a finite transitive clause in that it conceptually correspond to the archetype of an energetic interaction and structurally manifests the fullest array of properties that are specifically clausal. As a notion for the characterization of the finite clause prototypes, transitivity is seen not in terms of nominal occurring in a particular structure configuration, rather a matter of degree. In line with Hopper and Thompson (1980)’s work on the proprieties with reference to transitivity, Langacker believes that all those conceptual factors can be identified as facets of the canonical event model which tie them together and provide a basis for the prototypical notion of transitivity. For example, Jim broke my cup. The organization of this sentence reflects the canonical event model and it is quite high in transitivity because it corresponds to the prototype regarding multiple factors of the model. The nominal complements of a relational predication are often referred to as its arguments. The arguments of a prototypical transitive clause represent distinct, clearly delimited participants that are sharply differentiated from each other, from the profiled event, and from the countless other entities that could conceivably participate in their stead. Found in all languages, however, are grammatical devices that accommodate departures from this canon. Such devices permit the efficient coding of events in which the participants are not in fact distinct, or where the speaker is unable to specify an argument with any precision. The resulting expressions usually have diminished transitivity, even to the point of being fully intransitive (Langacker, 1991, p. 362).
The conception of two participants thus interacting is a central feature of the canonical event model, the prototype for transitive clauses. The participants of a prototypical clause are therefore separate and discrete physical objects that exist both prior to the profiled event and independently of its occurrence.
Our immediate interest lies with the broad class of non-conforming expressions in which a participant is not wholly independent of the designated process (Langacker, 1991, p. 362). A minimal lack of independence is exhibited by participants that have no prior existence but are rather created by the event itself: build a cabin; knit a sweater; carve a figure; paint a masterpiece; draw a diagram; light a fire; make dessert; cause a traffic jam; devise a plan; write a letter; compose a sonnet; strike a deal; create a consensus; invent a new technique; etc. Hopper refers to such participants as “effected” objects and surveys their properties in English and other languages. He finds that effected objects are seldom distinguished from “affected” objects by any special marking; they may however show behavior symptomatic of lesser transitivity. Observe, for instance, that He cut a slice of cake permits an effected-object interpretation whereby the cutting brings the slice into existence. However, adding the particle up imposes the affected-object construal: He cut up a slice of cake can only mean that a pre-existing slice was partitioned. Hopper notes that cut up represents a higher degree of transitivity than cut alone, for it generally indicates that the action affects the object in its entirety (e.g. He cut up the meat vs. He cut the meat).
2.2 Analysis of Langacker’s Argument on the Degree and the Meaning of Transitive Clauses in Transitivity A clause that has both a subject and a direct object is said to be transitive. In recent years, the traditional notion of transitivity has been investigated in considerable depth and has proved to be both subtle and complex (Hopper and Thompson 1980). There are cases where a verb is arguably transitive despite the lack of an overt object. For instance please in Roger always aims to please and dried in Peter washed the dishes and Marsha dried.
More seriously, the occurrence of a nominal in direct-object position does not invariably render a clause transitive, as determined by such traits (depending on the language) as case marking, agreement, passivizability, and special verb morphology characteristic of transitives. For example, although the sentences in (1) have post-verbal nominals that appear to be direct objects (on the basis of position and the absence of a preposition), the infelicity of the corresponding passives in (2) calls their transitivity into question.
(1) (a) After his divorce, Jason became a new person.
(b) That little boy resembles my nephew.
(c) Tuesday witnessed yet another gang-related slaying.
(d) The reactor’s cooling system burst a pipe.
(e) The former emperor died a peaceful death.
(2) (a) *After his divorce, a new person was become by Jason.
(b) *My nephew is resembled by that little boy.
(c) *Yet another gang-related slaying was witnessed by Tuesday.
(d) *A pipe was burst by the reactor’s cooling system.
(e) *A peaceful death was died by the former emperor.
Such considerations have made it increasingly apparent that transitivity is not definable just in terms of nominals occurring in a particular structural configuration. It is instead a matter of degree and depends on the meaning of the clause as a whole (Langacker, 1991, pp. 301-302).
Works such as Hopper and Thompson, Hopper, and Rice have shown that a substantial number of conceptual factors contribute to transitivity. Among the properties characteristic of a prototypical transitive clause are the following: (1) it has two participants expressed by overt nominals that function as subject and object; (2) it describes an event (as opposed to a static situation); (3) the event is energetic, relatively brief, and has a well-defined endpoint; (4) the subject and object represent discrete, highly individuated physical entities; (5) these entities already exist when the event occurs (i.e. they are not products of the event); (6) the subject and object are fully distinct and participate in a strongly asymmetrical relationship; (7) the subject’s participation is volitional, while that of the object is non-volitional; (8) the subject is the source of the energy, and the object is its target; (9) the object is totally affected by the action. This is not a random list. As pointed out by Rice, all these factors can be identified as facets of the canonical event model it is this model that ties them together and provides a coherent basis for the prototypical notion of transitivity. These sentences in (1) are quite low in transitivity because they deviate from the prototype in regard to multiple factors (Langacker, 1991, p. 302). While, In She likes it, for example, the profiled mental interaction (of the form EXPER→ ZERO) is construed as an abstract analog of the energetic interaction (AG→ PAT) that is prototypical for transitivity, so it is coded linguistically as a transitive clause (Langacker, 1999, p. 31). This we cannot afford to agree to. According to the above parlance, the sentences in (1) have post-verbal nominals that appear to be direct objects; the infelicity of the corresponding passives in (2) calls their transitivity into question, which is really arbitrary. Firstly, (1a) actually differentiates from the transitive verb resembles in (1b) since became itself has two lexemes; one is as transitive verb while the other is taken as the intransitive in traditional grammar. And a new person should be regarded as the intransitive whole, for a new person is a part of meaning of the intransitive and the complement of the complete meaning of the intransitive became. As for the intransitive verb, became indicates ‘come to be’ and as for the transitive one, it implies to ‘be appropriate or suitable to’, which shows the complex of the verb became. Therefore, we have to admit that the passive expression of a new person was become by Jason is eventually not reasonable semantically and is to be deleted, which is not contradictory to the infelicity of the corresponding passive in the transitivity problem without question. It is obvious that those listed examples are not representative in fact to reason the problem. Secondly, whether or not the passive pattern is overt or explicit poses no challenge to the transitivity. As a matter of fact the transitivity problem is the abstract category involving the aspects of language performances undoubtedly, but the transitivity is hierarchical and also recursive as the syntax deals well with this matter. And the absence of the passive form is not equal to the non-existence of covert structural computation. When it is manipulated by our mental operation, the appearance and dormancy can achieve a balance through struggling compromise or tuning up, that is to say, the result is the representation of the passive form or the vacancy, which is possible as well. There is a tension between the prominence and recessiveness, which reaches the higher unification by dint of the law of thought. If the practical phenomena take the place of the rank of reason, it brings out the chaos of the realization. The infelicity of the corresponding passives in (2) calls their transitivity into question finally has its view of the important overshadowed by the trivial, as we can not take the place of abstractness by means of the concreteness. The third, the infelicities of the corresponding passives in (2), all phenomena are merely the language representations since absence is another way of expression. And the correspondence between form and function or meaning is not the simple one-to-one relation, as our expressions are the shadow of the shadow of the idea or logos, in other words, our language performance is the replica of the language mechanism operation. Looking back at (1a), the verb became itself containing the transitive and the intransitive usages are the inspissations of the transitivity problem, and it shares the functions of the two poles by virtue of one form, that is, it is the isomorphic exhibit. We claim that the used or accepted passive transformation, or otherwise can not cancel the nature of the verb, that is to say, the transitivity problem is the category domain and the yes-no question, but irrelevant to the degree or the deviation from the prototype in regard to multiple factors. As the multiple factors themselves are impressed by the experiential and phenomenal operation and too much subjective, undoubtedly, they are not of the qualification of scientific rigidity and easy to collapse by means of few counterexamples. The details of falsification is similar to the latter, here we leap over. As to the voice of high or low in transitivity is ridiculous in practice, for the sensual, intuitional and psychological judgment or the data from the statistic operation is not persuasion whenever and wherever. After all, it is fluid or in flux and it is difficult to cater for all tastes, which is not rational, not persuasive and not convincing. Langacker has neglected other possible interpretations of an expression. We combine words into sentences with pre-determined lexical meanings of each word, and know which meaning is intended. If it demands, the null or empty elements could be complemented through our construal process and understanding. He takes radial instead of hierarchical relation. Further, there is no ‘surplus’ usage accruing to a verb no matter transitive verb or intransitive one that is not attributable to its principle and parameters mechanism. Universals may be breached but they need not be present. It is stated that the rules of language do not consider simple linear order but are structure-dependent. Structure-dependency is a discovery about the nature of human language in general, and a principle of Universal Grammar.
What distinguishes transitive verbs from intransitive ones is the syntactical structure instead of semantic values. And passivity cannot be used as the only standard of the transitives. Since passivization is a manipulation and one sort of representational expressions in transitivity, which makes active sentences such as Tom killed him into passive sentence He was killed by Tom. In this process, the object of the active sentence is turned into the subject of the passive sentence. The two sentences are different in their respective viewpoints. As to the above two sentences, in the active aspect, its viewpoint is neutral, on the other hand, in the passive sentence the speaker describes the event only from the patient’s view. However, not all the events or states expressed in transitive sentences can be described from the patient’s point of view. This is the matter of perspective or construe in humans’ eyes.
In this case, it is construed and checked as the erasing or absence of passive usage. However, the desuetude in this situation does not mean the non-existence of the concealable language mechanism. It is pivotal that grammar is there. It works on its own and with resources from relexification, which has reified through interaction between language faculty and language use (Zhao Yanchun, 2006, p. 19). In communication, we often omit many elements of the utterances as default values like the given example dried in Peter washed the dishes and Marsha dried which has been mentioned at the very beginning of this chapter, but have no reason to veto the transitive nature of the verb dried, as its indicated object is covert or not necessary to appear in virtue of the relevant and economic principle of expressions. Again rethink the mentioned examples Jerry bumped into John and John was bumped into by Jerry, which are mentioned in chapter two already. Through the further syntactic analysis, we regard that was bumped into itself is a verb phrase and need not constituent-select or category-select the object as a norminalized phrase, and it is the logical objected. And it is the verb phrase was bumped into that semantic-selects the theme John, not the patient. Therefore, through the reference of that deep syntax analyzing, we point out that syntax has no semantic meaning or no semantics concern and syntax is autonomous, which can elaborate the passivity matter and give a well done.
And in The Norman defeated the Saxons, the transitive verb defeated category-selects an NP the Saxons, while semantic-selects two θ roles, one is the agent The Norman and the other is the patient the Saxons. In the same way, in The Saxons were defeated, were defeated as one node; that is to say, one verb phrase has no category-selection and semantic-selects one θ role The Saxons as the theme. Through the syntactic computation, the passivization phenomena can be made clear. It is stated that phrases are members of syntactic configuration, and logical form, partial representation of meaning is determined by grammatical structure. In syntax, phrase grammar is rather than word grammar as were defeated is taken as a constituent in syntactic operation. The links between syntax and the lexicon should be definite and important because structure-dependency also affects the interpretation of sentences. Lexical structure must be represented categorically at every syntactic level. The transitivity problem can be explicit by the way of syntactic modular operation of language mechanism.
From another perspective, meaning might be the basis and warrant of understanding classification to some extent, but not the essential or the fundamental level compared with the grammatical structure’s criterion. Dividing into transitive verbs and intransitive verbs serves syntax analysis, so the standard of classification should be the formal one, that is, objects can follow the transitive verbs directly while the intransitive verbs cannot in the semantic way. The coming question is what the objects are, that is to say, what elements can be used as the criterion of transitivity’s ranking. It is worthy to use for the reference differentiating the complement or not in the formal valence grammar. When deleting some elements, whether the left sentence structure accords with grammar or not, if it is answering for it, the erasing components may have its complement or free expository points; if not, the erasing is to be followed by the complement. In so many words, the formal valence view for distinguishing transitive and intransitive verbs seems an undoubtedly newly approach. It is clear that semantics must be distinguished from syntax with no strings attached, since they are two different modules governing their own operational ranges. On occasion, semantic meaning may be neglected when it is not too necessary, and it is covert or implicit. As syntactic structure is not indispensably determined by semantic meaning but able to make clear the semantic value, we can claim that syntax is self-independent and autonomous. Such as love, the expressed same semantic meaning might be a noun or a verb as well in syntax, which discloses its kind that is determined by the syntactic structure, irrelevant to the semantic aspect. The following (3) can present a clear explanation.
(3) a. I love him.
b. My love for him is a mistake.
c. Love is great.
In (3), love in all those three sentences shares the same semantic meaning undoubtedly, while the distinct class features that (3b) and (3c) compared to (3a) possess, the former two belong to the noun category and the latter the verb in the totally different syntactic structures from the syntax perspective. Since in syntax the structure-dependency determines the various syntactic features of love, irrelevant to the meaning, it is stated that semantics and syntax are different modules and syntax is autonomous without question.
Returning to Langacker’s evidences, we have our rational thought. As the case stands, the infelicity of the corresponding passives in (2), that is to say the forms of (2) make some information default, which is not the comprehensive problem or the semantic understanding to the quick, but the matter of non-conventionalization or the result of non-phylogeny, that is, they are not taken in for the moment, which does not mean they bear no potential competence in syntax.
Furthermore there is a tension between transitives and intransitives, and the minimal pair of them exists. Any expression is the representation of the structure-independency in language; therefore, all the absent information of (2) shows the recessive application of transitives in (1).
As to the displayed hypotheses, are all those nine factors really all-sided to be as principles or the mode of checks and balances in transitivity? However, it is the fact that it will have our view of the importance overshadowed by the trivial in conformity to their asseveration unfortunately. Firstly, it is based on the experiential representation of language, not satisfying stringency. Therefore, meaning is involved to analyze and induct the transitivity, which is inadequate to the operation of transitivity. Since semantics and syntax are independent of each other, the former is the expressional system of language meaning and the latter is the computational system of language mechanism, and there is no absolute and necessary correspondence relation. The second, its depth of field is rather confined by limited expressional resources, if it sounds with reason, which is considered as a local availability, because in fact there exist a large number of disproval exemplifications to refute those statements. It points out those properties characteristic of a prototypical transitive clause collapse in the face of indelible convincing counterexamples. Thirdly, similar to the above nine factors, Taylor also fetches a series of similar hypotheses. We adopt the logic tool to challenge Langacker predication by falsification and also Taylor’s collapses of itself finally. 3 Discussion of Taylor’s Assertion of Typical Instantiations
3.1 Taylor’s Arguments
Taylor (2001, p. 206) brings forward the prototypical hypothesis in transitivity. The followings are typical instantiations of the transitive sentence construction.
(4) The child kicked the ball.
(5) John moved the table.
(6) Mary killed the intruder.
The syntactic properties of the construction may be represented by the formula NP1VTRANSNP2, where NP1 and NP2 stand for the subject and direct object, and VTRANS is a transitive verb. In its prototypical instantiations, both NPS have specific reference, while the verb is realis, i.e. affirmative and indicative, and in a reporting tense (either present or past). Drawing on Lakoff (1977) and Hopper and Thompson (1980), we can list at least eleven semantic properties of the construction, in its prototypical instantiations.
(a) The construction describes events involving two, and only two participants, encoded by the subject and direct object NPS respectively.
(b) The two participants are highly individuated, i.e. They are discrete, specific entities (from this it follows that both the NPs in the construction have specific reference), distinct both from each other, and from the background environment.
(c) The event is initiated by the referent for the subject NP, i.e. by the agent. Responsibility for the event thus lies exclusively with the agent. Furthermore, the subject NP is the sentence topic; the subject is what the sentence is about.
(d) The agent acts consciously and volitionally, and thus controls the event. Since consciousness and volition are typical human attributes, it follows that the agent is typically a human being.
(e) As a consequence of the agent’s action, something happens to the patient, i.e. the referent of the object NP. The effect on the patient is intended by the agent. Often, though by no means necessarily, the patient is inanimate.
(f) After the occurrence of the event, the patient is in a different state from before the event. Usually, the difference is one which would highly perceptible to an overlooking observer.
(g) The event is construed as punctual. Even though the event necessarily has temporal extension, the internal structure of the event, and the intermediate states between its inception and termination, are not in focus.
(h) The agent’s action on the patient usually involves direct physical contact, and the effect on the patient is immediate. (i) The event has a causative component —the agent’s action causes the patient to undergo a change.
(j) Typically, agent and patient are not only clearly differentiated entities; often they also stand in an adversative relationship.
(k) Finally, the events reported by the construction are real, not imaginary, hypothetical, or counterfactual. Hence, central instantiations of the construction are realis.
(Taylor, 2001, p. 207)
3.2 Analysis of Taylor’s Arguments
Let us see whether Taylor’s view is tenable. Now we shall analyze Taylor’s arguments length by length. Let us use Taylor’s standpoints for our counter-argument.
Taylor regards that prototypical instantiations of the described events involve two, and only two participants, which is not reasonable actually according to plentiful basic language facts, like She gave me a book. In this case, obviously there is not merely a simple question of the subject and the direct object. Nevertheless, we know quite well that one direct object and one indirect object follow the transitive verb gave. Therefore, (a) is not taken as the respective criterion of the prototypical transitivity.
As for (b), again Taylor has overlooked the language fact, such as You should help each other and I hate myself which exists everywhere. Considering that fact of language, nobody I suppose can agree with Taylor, for you and each other or I and myself are the same reference or entity, that is, the latter each other or myself refers to the former you or I.
As regards statement (c), which asserts that the agent is responsible for the whole event, we agree that the process of the event absolutely connects with the subject, however, he says the subject NP is the sentence topic is not persuasive for every aspect.
Explicitly, Taylor’s fallacy is that he has over-generated the entry to transitivity following (c). For instance, the expression Happiness gave me a lot of inspiration is a reasonable and acceptable transitive clause. However, inanimate Happiness is taken as the subject, i.e. the agent, which is rather legal after all.
Taylor’s argument is so subjective that it is full of contradictions. Again look back at the former example Happiness gave me a lot of inspiration. With regard of this case, me is one of the patients logically and an indirect object and a human being without question, which is rightly in contradiction with the assertion that the patient is inanimate. With regard of this case, Taylor’s argument seems explicit; however, he mixes the different levels of language. The operational mechanism and the understanding realizations are not the same process. Again like I hate myself, the internal change is not grasped by means of the representation of language, and the interpretations of the understanding process are various.
Taylor’s argument is too arbitrary and unilateral in (h). Firstly, the sayings ‘direct physical contact’ is rather limited and insufficient as the agent’s action, which deals with the psychological and mental recognition, considering the effect is quite changeable according to different people and different situations. Secondly, The effect on the patient is immediate is the only tip of real facts. Look at the following example I see what you mean, there is no direct physical contact and the effect is involved in the patient, not as simple as the immediate reaction after all.
As for (i), what kind of mistake Taylor makes is that he comes to conclude through the limited experience. How to confirm that the agent’s action causes the patient to undergo a change is reasonable, which may be really hard. For instance, in I miss you very much, what change is implied, it is not easy to answer.
In the same way, as to (j), we can find at least one exception to falsify Taylor’s arguments. Since agent and patient are not clearly differentiated entities, their relationship is not sure to be only the case of adversativeness. In fact, the relationship is not simple just as what he says, but is plentiful according to colorful language resources. Such as in I enjoy myself, myself is the anaphor of I, so they are tied by the same relationship.
In (k), Taylor takes his every effort to maintain his opinion of prototypical transitivity, however, he overlooks the mechanism of language’s syntactic operation and the law of human’s logic thought. He overgeneralizes and draws a conclusion of that so-called central instantiations.
More fundamentally, Taylor’s problem as well as that of other cognitivists lies in the negligence of the nature and/or mechanism of the human speech, which results in the inaccuracy of taking the performances or representations of language for granted as the maxims of the transitivity entry. Taylor’s claim of transitivity has theoretic weakness resulting from their abandoning of the classical category theory. Now that the above detailed rules cannot be verified, but to be falsified. 4 A Non-Prototypical Transitivity Verdict
In short, the above linguistic evidence stringing along with their positions pose no threat to the classical account of transitivity. If applications are taken as the transitivity category itself, our lexicon would not have such a capacity, and communication would be impossible. This problem has been elaborated by Aristotle. Evidently, the above hypotheses are not necessary.
In light of their ‘It goes without saying’, the prototypical transitive clause possesses most of the semantic properties, thus most representative of transitivity category. It is evident that all of these semantic features have their experiential basis, corresponding to Langacker’s canonical event model. What is more, they believe that this fact also coincides with the interactional nature of properties that are seen to determine the prototype of a category, and such bodily interaction with the outside world Lakoff believes from clusters in our experience, and prototype can reflect such clustering. However, how to judge the highest level of transitivity in terms of those rules?
We have to point out their declarations resting on physical, psychological, and mental refraction, which cannot hit the core of problem. Although, relying on realistic embodiment, we really construe the world and achieve our realizations to some extent, which merely acts in the superficial stage whatever happens. Since the world is not what we have seen, language is representational, and category is abstract; this eventually educates and leads us to go deep into the nature of phenomenon instead of concrete and experiential things. Thus problems are coming to us. Is the determination of prototype definite and clear? Are transitives prototypical and the pre-appearing to be undoubtedly prototypical as well? Then which one is the pre-existence, transitives or intransitives? Is it possible that as long as it is more applying frequency, then more prototypical? Frankly speaking, prototype category is not able to answer these fundamental questions of transitivity problem, and then the prototypical transitivity is on the verge of imminent death. Therefore, the so-called prototypical transitivity wherewith to shake the ground of the foundation of the classical category is but a whim. Considering the analysis of Langacker’s arguments and Taylor’s prototypical hypotheses in transitivity cannot hold, the non-prototypical transitivity should be affirmed ultimately.
[References]
[1]Hopper, P. & Thompson, S.A. Transitivity in grammar and discourse. Language. (1980), 56.
[2]Lakoff,G. Linguistic gestalts. In CLS (13), Papers from the 13th Regional Meetings. Chicago: Chicago Linguistic Society. 1977.
[3]Langacker, R. W. Foundations of cognitive grammar (vol. 2): descriptive application.. Stanford: Stanford University Press. 1991.
[4]Taylor, J. R. Linguistic categorization: prototypes in linguistic theory. Beijing: Foreign Language Teaching and Research Press. 2001.
[5]赵彦春.语言哲学(课程).2006.
本文系2012年度中央高校基本科研业务费专项资金资助(supported by the Fundamental Research Funds for the Central Universities)“及物性原型观的批判性研究”(2012RW016)
(作者单位:华中农业大学外国语学院, 湖北 武汉 430070)