Thepkanjan, Kingkarn (2000) “Lexical Causatives in Thai” in Constructions in Cognitive Linguistics: Selected Papers from Fifth International Cognitive Linguistics Conference, Amsterdam 1997 (eds.) Ad Foolen & Frederike van der Leek: 259-281. Amsterdam/Philadelphia: John Benjamin Publishing Company.

1. Introduction

causative constructions can be classified into the major types based on the productivity or regularity of causative forms, namely, prodcutive causative and lexical causatives. – 259

Productive causatives are realized by the use of causative verbs, such as English cause and make and Korean ha-ta, or the use of affixes, such as Japanese sase. This means that productive causatives may be further classified into two subtypes: syntactic and morphological causatives. – 259

Syntactic causatives, which are alternatively called analytic causatives or periphrastic causatives, are generally defined as the causatives formed by means of specific verb of causation such as make, have, cause in English, fare “make”, lasciare “let, allow” in Italian, and hay “let, allow” in Thai. -259

Morrphological causatives, alternatively clled synthetic causatives, on the other hand, are defined as causatives that are realized by means of morphological devices applied to verbal forms, such as affixation in Russian and vowel alternation in Armenian. -259

Lexical causatives refer to morphologically irregular, nonproductive causative froms. Lexical causatives are typically manifested in languages as a class of transitive verbs referred to as causative transitive verbs, such as cut, destroy, open, melt, kill and boil in English. -259

It could be claimed that a lexical causative which is manifested in the form of a transitive verb, expresses two major subevents: the agent’s activity and the patient’s change of state. – 259

2. The windowing of attention in language:

The windowing of attention in language is defined as a cognitive process which places a portion of an event-frame into the foreground of attention by explicitly mentioning it, while placing the remainder of event-frame into the background of attention by omitting mentioning it. The portions that are foregrounded are said to be ‘windowed’ and the ones that are backgrounded are said to be ‘gapped’. – 261

According to Talmy, the windowing of attention is only one part of the larger cognitive structural category in language called the ‘distribution of attention’, which is a system constituting the fundamental delineation of conceptual strucutring in language. – 261

3. Semantic differences between synatctic causatives and lexical causatives

The syntactic causatives, which consist of a two-verb structure, names two separate events whereas lexical causative expresses one single event with two subevents constituting the causative situation (Shibatani 1976).  – 262

The crucial point is that the cause and the effect in the syntactic causative are felt to be two separate events whereas those in lexical causative are percieved as jointly representing one single event. Haiman (1985) attributes this difference to the role of iconicity in language. He claims that the linguistic distance in any structure is generally motivated by the conceptual distance. – 262

Haiman claims that the syntactic causative with two separate verbs has more linguistic distance than the lexical causative and thus signals more conceptual distance between causer and causee. – 262

Prototypical causation is understood in terms of a cluster of ten interactional prperties, such as the presence of a human agent who does something, the presence of a patient which undergoes a change of state, the human nature of the agent, the volition of the agent, an overlap in time and space of the agent’s action and the patient’s change of state, a direct contact between the agent and the patient, etc. – 262

The type of causation that has the whole cluster of ten properties is categorized by Lakoff as prototypical causation. Nonprototypical varities of causation are characterized in terms of deviations from the cluster. – 262-263

Lexical causatives tend to signal direct, manipulative causation whereas syntactic causatives tend to express indirect causation. – 263

In direct causation, the effect is brought about directly by means of direct contact by the causer’s action whereas in indirect causation the effect is brought about through a mediating agent. – 263

According to Lakoff (1987), direct causation characterizes prototypical causation whereas indirect causation is the non-prototypical kind. – 263

Lexical causatives do not necessarily express direct causation in the strictly objective sense, such as in the case of the verbs build in I will build a house made of teak and clean in I cleaned this suit at the dry-cleaning store. – 263

Talmy (1996) explains this phenomena in terms of the theory of windowing of attention in language. […] the windowing of attention in language is defined as “the system with which languages can place a portion of a coherent referent situation into the foreground of attention by the explicit mention of that portion while placing the reminder of that situation into the background of attention by omitting mention of it.” (Talmy 1996: 235) – 263-264

The cognitive process in which the medial portion of the casual chain is reduced in conscious conceptualization to the degree tha the discontinuous initials and final phases may seem to be seamlessly linked to each other, is termed ‘conceptual slicing’ by Talmy (1996: 249) – 264

Conventionalized scenarios “can be cognitively packaged as a single event even if an interesting cause exists.” (Goldberg 1995: 169) - 264

4. Types of lexical causatives in Thai

If we adopt the definition of a lexical causative as a transitive verb which designates a situation involving an activity representing a cause and a subsequent change of state representing the effect, the lexical causative will, […], count as an accomplishment verb as defined by Dowty (1979). – 264

4.1 Alternating transitive verbs

4.1.1 Transitive causative verbs alternating with activity verbs

4.1.2 Transitive causative verbs alternating with inchoative and stative verbs

A clausal chain according to Croft (1991) is a cognitive model of conceptualizing events in the world, which is based on the interactions between entities and the transmission of energy in the interactions. The clausal chain is thereore used as a cognitive model to represent causation. – 266

4.1.3 Transitive causative verbs alternating with stative verbs

4.1.4 Semantic properties of alternating transitive verbs

The agent could move the car by driving, In case the car moved downhill, the agent could push the car for a few moments, stop, and let the car move downhill by itself. The former case exemplifies the entraining causation type whereas the latter one exemplifies the launching causation type. – 271

4.2 Non-alternating transitive verbs

4.2.1 Non-alternating transitive verbs with inherent effects

A telic situation refers to a situation which consists of an activity which leaps up to a culmination point, beyond which the activity cannnot continue. An atelic situation, on the other hand, refers to a situation which is realized as soon as it begins; it does not have to wait for a goal to be attained. – 272

According to the theory of aspectology, the telicness of situations can be designated either by verbs alone or by verbs combined with complements and/or modifiers. In the latter case, the telicness is determined by the interactions of several factors such as the verbs inherent meaning, and determiners, and the nature of the verb’s arguments, that is, of the subject and object(s). – 272-73

The imperfective aspect “denotes a complete situation, with beginning, middle, and end” (Comrie 1976: 18). The perfective aspect indicates a situation which is in progress and which lacks an inceptive stage and/or a terminal stage. – 274

4.2.2 Non-alternating transitive verbs with potential effects

4.2.3 Semantic properties of non-alternating transitive verbs

5. The so-called suppletive lexical causative forms

6. Conclusine

Takashi, Hidemitsu (2000) “English Imperatives and Passives” in Constructions in Cognitive Linguistics: Selected Papers from Fifth International Cognitive Linguistics Conference, Amsterdam 1997 (eds.) Ad Foolen & Frederike van der Leek: 239-258. Amsterdam/Philadelphia: John Benjamin Publishing Company.

1. Introduction

Early generative grammar work claimed that English imperatives disallow the passive forms syntactically (cf. Lees 1964), based on facts like those in (*Be helped by Jill) above. However, later pragmatic and descriptive work found that this restriction is far from absolute (cf. Bolinger 1997; Davies 1986). – 239

2. English imperatives and their four features

In English, the ‘imperative’ refers to a construction which occurs only in main clause, normally has no grammatical subject and contains the verb in its basic form. – 240

[Takahashi] characterized English imperatives in terms of four essential features, hypotheticality, non-past, second person and force exertion. – 240

Force Exertion: the degree of (directive) force that the speaker is applying (at the utterance time of an imperative) toward the addressee’s carrying out the action. – 240

3. A cognitive (image-schema) model of imperatives

3.1 Action chain model

i. The imperative is comprised of two separate (though interrelated) subevents, which combine to from a dynamic chain of action;

ii. The speaker and the addressee both participate in this action chain as indispensable entities – the speaker as head, the addressee as a second entity engaged in further action;

iii. The imperative makes explicit only the addressee’s action and leaves implicit the two key entities; nor does it overtly code the application of directive force. – 242

3.2 Canonical event model

The simplest transitive clause contains two participants which play semantic roles, agent and patient, respectively. These participants are engaged in a kind of energetic interaction in a specific conceptual domain called a ‘setting’ chosen to be highlighted by a conceptualizer. A transitive clause such as John ate the apple in the kitchen, where John acts as an agent and participates in an energetic interaction of eating an apple which is patient in the ‘kitchen’ setting. – 243

To obtain a Canonical Event Model of the imperative, let us specify the setting and the semantic roles of participants. As for the setting, the first subevent is restricted to an extremely narrow speech situation, the here-and-now of speaking, only populated by the speaker and addressee. It would be reasonable then to analyze this subevent as occurring in a deictic setting. On the other hand, the second subevent (or the addressee’s action) is hypothetical is native, regardless of whether or not the addressee’s act will be reanalyzed in the objective world. For this reason, the second subevent may be treated as taking place in a hypothetical setting (or mental space). – 243

The speaker can be analyzed in terms of cause-agent, and the addressee in terms of cause-agent. The rationale behind this treatment is that in the prototypical scene of an imperative, the addressee is triggered to act or undergo a substantial change in state by the utterance of an imperative – 244

The prototype IMPERATIVE

i. The speakers exerts a high (near [+1]) directive force in deictic setting towards the addressee, who will thereby perform an action in hypothetical setting.

ii. The speaker plays a semantic role of causer-like agent, and the addressee cause-like agent. – 244

The schematic IMPERATIVE

i. The speaker exerts a varying degree of force (ranging from [+1] to [-1] in deictic setting toward the addressee, who will thereby be engaged in a certain situation in hypothetical setting.

ii. The speaker and addressee may play an agentive as well as non-agentive role. – 245

(A) Imperatives are atypical when the speaker’s force in non-directive (or mixed with other kinds of force). […] (B) When the subject is a non-agent such as an experiencer, theme or patient or when the understood subject is generic as in Shake before using, the imperative progressively deviates from its norm. (C) Events (highly dynamic situations) better fit the image schema of prototype imperatives than states as in Be sick (and they’ll put you in bed). - 246

4. English imperatives and passives

It is well established that an active clause more readily passivies if it is more transitive, i.e. when the patient is directly affected by the activity in question. – 247

The passive critically involves an agent (or something close to it such as an experiencer), which is invariably defocused either syntactically or lexically. Second, the passive with be is in principle more static than its active counterpart, alhtough this does not mean that the be-passive only refers to a state. Third, the subject tends to be a patient. – 247

The prototype passive

i. The subject is directly affected by an external agent.

ii.  The subject plays the semantic role of patient. – 247

The subject’s patienthood, the defocused participant’s agency and overall affectedness normally go hand in hand. One almost automatically follows from another, since prototypical patient is a participant absorbing the energy transmitted from without and thereby undergoing a change in state. – 248

Hypothesis: The passive construction does not clash with the imperative syntactically but on conceptual grounds, i.e., the clash occurs between prototypes. – 249

5. Japanese imperatives and passives