An analysis of the Q–A unities constructed by the interrogative pronoun какой

1 Introduction

According to Saussure, language is a system of signs that develops over time. There is an asymmetrical binary correspondence between the signifier and the signified. This phenomenon is also widespread in Russian linguistics. Russian belongs to the East Slavic group of the Slavic branch of the Indo-European language family. In its morphological structure, Russian is a predominantly inflective, synthetic language (Karaulov 1997, pp. 437–444). Russian lexicography holds that the interrogative pronoun какой asks about a certain quality or a certain attribute (USSR Academy of Sciences 1956, p. 693; Ozhegov and Shvedova 2006, p. 260). In addition, some dictionaries indicate that this interrogative pronoun can be used in spoken language to ask order (Shvedova 2002, vol. I, p. 30), such as the order of choice, the order of arrangement, or the temporal order in which things appear (Kuznetsov 2000, p. 410). The Russian Grammar (1980) categorizes it as an interrogative adjective subordinated to the adjective category and states that this interrogative pronoun can be used to ask time, time amount, space, cause, purpose, condition, distinction-restriction, manner, quality, quantity, or other aspects (USSR Academy of Sciences 1980, vol. 2, pp. 160–162). However, as Kreidlin and Rakhilina (1984) pointed out, “this incomplete semantic representation of the interrogative pronoun какой in questions lies in, to a large degree, the fact that the studies on it are conducted outside of the conversational unity.” That is, some answers may satisfy the semantic requirements established in dictionaries about the interrogative pronoun какой, but such answers are not acceptable, for example:

Analyzing the question itself in isolation would lead to an incomplete understanding of the interrogative pronoun and a failure to fully understand its characteristics.

The какой-question is also fundamentally different from other wh-interrogative sentences. In interrogative sentences containing interrogatives such as где (where), кто (who), когда (when), etc., the interrogative variant is closely related to its relevant semantic valency; hence such an interrogative is associated with a single semantic argument. In contrast, in the какой-question, какой is associated with several arguments, which the respondents can address freely and selectively, filling in multiple valencies. In view of this, we propose to study the какой- question in the context of the Q–A unity. The Q–A unity is a relationship of correspondence or consistency between the interrogative sentence and its reply. Analyzing the features that distinguish it from other wh-interrogative sentences, we try to clarify the understanding of this interrogative and fill the gaps in its traditional lexicographical interpretation.

2 Types of какой questions

According to the morphological-syntactic characteristics of какой questions, they can be divided into two categories,^[1] which can be expressed symbolically as (Ⅰ) Какой W Р? and (Ⅱ) Какой S ∅ N? ^[2] The first type is more complex in the construction of Q–A unities.

In this construction, W refers to the noun or noun phrase while P the predicative. Grammatically какой is subordinate to W, but semantically W is subordinate to какой. That is, such interrogative sentences ask questions about W. From the perspective of logic, the answer to such interrogative sentences is in fact a choice of the object of the W set. The interrogative variant is the constructive part of the logical form of the questions, and linguists equate it with the semantic representations of the interrogative sentences.

In this construction, S is a noun or pronoun functioning as a subject, ∅ is a zero-predicate, and N is a noun in the predicate phrase. When S is a noun, the above-mentioned two structures though different in grammatical nature can lead to the same syntactic form; in other words, the same interrogative sentence may be interpreted not only as structure (I) but also as structure (II), for example,

3 The syntactic-semantic relations of the Q–A unity constructed by the interrogative какой

When constructing a qualified Q–A unity for such interrogative sentences, we need to take into consideration the grammatical properties of W.

If W is a gerund, the answer is usually the meaning of the variable term when it is interpreted, as the question mentioned above:

We can answer like this: – Раскопки Кремля (– The excavations of the Kremlin), because the object argument of W in the question here is incomplete and it is supplemented in the answer. Answers in similar contexts are often not effective, straightforward or even difficult to accept if they follow the dictionary’s interpretations of какой in terms of quality or attribute.

Another typical example given by Rakhilina:

In this Q–A unity, in response to the gerund салют in the question, all the arguments of this gerund appear in the answer: кто (who), кому (to whom), за что (for what), чем (how), из чего (from what).

It shows that the common practice when constructing the correct Q–A unity for such какой interrogative sentence is to fill in the free semantic valency of that gerund. Rakhilina summarizes its logical-semantic formula as ?Х<{Х | W (X)}P(W)>, in which X is the interrogative variant used to complement the valency of the gerund W (Rakhilina 1990). In principle, when answering this type of question, the respondent can provide several or even all arguments. Nevertheless, usually only one or two arguments will appear in the answer. Moreover, the priority of these arguments varies, with the content argument appearing the most frequently and the subject argument appearing the least. However, this order of priority does not prevent the subject argument from appearing in answers in the context as:

The word проект in the question is bivalent, i.e. “who designed it” and “what was designed”, and the content argument is already specified in the question, so it is necessary to point out the subject argument in the answer.

If W is not a gerund, then answers about characteristics, properties or nature is not always qualified, e.g.

In similar contexts, it is preferred to indicate the proper name in the response. There is another situation that needs attention. That is, the content of the answer should be consistent with the object domain that the question refers to; if not, the answer will not be what is being asked. Here is an example given by E. V. Paducheva:

The interrogative object domain is posed by the interrogative component and its degree of saliency varies. The interrogative sentence constructed by какой is characterized by the polysemy of the object domain of the question, which is caused first and foremost at the syntactic level. For example, the question Какие чувства наполнили душу Ибрагима (What feelings filled Ibrahim’s soul?) can be understood in two ways, (a) the interrogative component is какие чувства (what feelings), and (b) the interrogative component is какие (what kind of feelings). In the first understanding, the interrogative object domain is the set of emotions (jealousy, fury, despair, etc.) that may occur, and in the second, the interrogative object domain corresponds to the set of attributes of the emotions (noble, base, violent, etc.). The polysemy of the interrogative component may cause communicative failure, such as,

The interrogative component referred to by the questioner in this situation should first be какой анекдот (what joke), i.e. the questioner wants an answer about the joke itself, while the respondent understands the interrogative component as какой and answers only its attributes.

In some cases, when W is not a gerund, it is not clear that the answer is derived from the valency structure of the word. At this point, it is possible that the answer comes from its paraphrased component. Rakhilina explains the word аппетит (appetite) as “способность и желание У-а принимать пищу (Y’s ability and willingness to eat)”, and a respondent may answer the question Какой у вас аппетит (What’s your appetite?)? with отличный (very good), превосходный (excellent), так себе (just so-so), etc. The two predicates способность (ability) and желание (willingness) in the interpretation have a “parametric” valency (Apresyan 1974, p. 64), so the answer can be paraphrased like this:

Thus, the valency of the predicate used to explain the word W can be filled in by the interrogative variant of the relevant какой question.

Encyclopedic information relating to S is important when we answer this type of structure. This is clearly illustrated by the following examples:

This structure has a number of variants in actual discourse; specifically, there are two varieties. Firstly, the какой N used as a predicate element appears in the form of indirect case:

Secondly, the zero predicate tense is tagged with temporal markers:

4 Limitations of the Q–A unity constructed by какой

From the above analysis, we can see that for the first type of Какой W Р? structure, the following conditions are required for the construction of its Q–A unity:

1. The presupposition of such an interrogative sentence is that a certain set of objects must exist, called the interrogative object domain by Paducheva (предметная область вопроса, ПОВ);^[3]

2. The denotative status (денотативный статус)^[4] of the noun in the answer should be consistent with the denotative status of the noun in the subject position (W) in the question.

Thus, in the case of condition (a), if a certain set of objects for that type of interrogative sentence does not exist, or if object selection in a certain set is not possible, then that type of interrogative sentence cannot be used, for example:

As for condition (b), in general terms, it means that the respondent should not use different expressions for the same object or the same expression for different objects when answering, for example:

The reason for the failure of this type of Q–A unity is that although it may be uncertain to the questioner about книги (books), it also contains denotative meaning, whereas in terms of the answer, некоторые (some), which denotes existence, and всякие (every), which denotes universality, are noun phrases that do not contain denotative meaning. Thus, expressions with a denotative meaning and expressions without a denotative meaning cannot be put together.

The following example is a better illustration of the situation. The respondent can answer the question Какое животное самое выносливое (Which animal is the most hardy?)? with (a) Наш слон Соня (Our elephant Sonia), or with (b) Ишак (Ишаки) (The donkey(s)). When answering with (a), the respondent’s understanding of животное (animal) is specific and referential, and is present in the common background knowledge of both parties. When answering with (b), the respondent understands животное (animal) in a non-referential sense and the answer has the referential status of a category attribute. Thus, if the questioner is specific and referential about животное (animal), while the respondent answers with an expression of the category attribute, then it does not constitute a qualified Q–A unity, and vice versa.

In addition, if W is not a gerund, the replacement of the interrogative variant needs to be considered in terms of the following three aspects of the W:

1. The domain of compulsory syntactics (зона обязательной синтактики): if Какой W refers to a proper name, then the respondent should, in most cases, point out that proper name, e.g.

The formal-logical representation of such interrogative sentences is ?X < {Х|Х∈Nom(W)} P(W)>, in which Nom (W) denotes the set of meanings of the variables of the word W in the domain of compulsory syntactics. In some contexts, numerals can also be used as proper names, as they can also specify individuals in a set. However, there are still differences between numerals and proper names. Numerals must be within a finite set in order to have a definite meaning. Furthermore, there are finite sets in which the objects are ordered, such as the houses in the street, and some finite sets in which the objects are disordered, such as the trees in the forest. Thus, whether or not the object set of W is structured in some closed set determines whether or not substitution expressions can be made in the response with numerals. For example, we know that floors are ordered and this order is usually expressed with the help of ordinal numerals. In Russia, houses can be marked with compound numerals like 2/3. Thus, when the set of objects of W has a closed structured order, the numerals, like the proper names (the names of a person, objects etc.), can be considered as a means of filling the domain of compulsory syntactics.^[5]

1. The domain of lexical functions (зона лексических функций):^[6] some Ws contain several subcategories, and this information is important to the Q–A unity, because the lexical-functional meanings associated with the W can be used to answer какой interrogative sentences with the formula expressed as ?X <{Х|Х∈Spec(W)} P(W)>, for example,

Some answers may be difficult to distinguish at first glance whether they belong to (1) or (2), for example,

It depends whether the answer is considered to indicate the proper name (1) or the category (2). However, it is important to note that the subcategory information does not always coincide with the name of the object, for example: the subcategory of ‘celestial bodies’ - planets, the proper name – Earth; the category of literary works – novel, title of a work – The Master and Margarita.

1. The domain of encyclopedic information (зона энциклопедической информации): sometimes the answers are related to the encyclopedic information contained in the W. This kind of information may be either general, such as the information for all nouns on size, shape, color, etc., or narrow, only for certain categories of words, such as the material of an object, the character of a person, etc. The formula of asking for encyclopedic information related to W can be expressed as ?X <{Х|Х∈U ⁱ _j } P(W)>, in which Uⁱ stands for the encyclopedic information of W, and U_j is the set of meanings corresponding to the semantic variables in the domain of encyclopedic information, such as the colors, the size and the materials, etc. For example,

For the second type of structure Какой S ∅ N?, the conditions for answering it are much less restrictive. When answering interrogative sentences of this type of structure, in many cases, as traditionally understood in lexicography, it is possible to state the qualities, attributes, etc. associated with S, for example,

In some contexts, the above requirements of the domain of compulsory syntactics also apply to this structure, for example,

In this situation, undoubtedly, the name of the city takes precedence in the priority of the information when the respondent provides information to the questioner that he or she does not have.

5 Conclusion

The above analysis shows that the traditional lexicographical and grammatical understanding of какой and the interrogative sentences constructed with it has certain limitations and does not fully reveal the characteristics of какой interrogative sentences. Therefore, we analyze this syntactic phenomenon by including it in the Q–A unity. Based on the above discussion, when answering какой interrogative sentences and constructing Q–A unities, we should understand them in terms of argument structures, generic relations, deixis, relevant characteristics, properties, or attributes.

At the same time, the rich semantic content of interrogative sentences constructed with какой requires us to avoid analyzing them in two ways: firstly, to formalize them in a one-sided way and to reduce the mechanism for generating answers to syntactic substitution, which would seriously narrow the category of correct answers, since substitution is based on the repetition of the syntactic structure of the question by the respondent. If the respondent does not repeat the syntactic structure of the question, then the substitution method is useless. Such an analysis would be divorced from the actual Q–A context and would be isolated from the Q–A unity. The second way that we need to avoid is to logicize it by reducing the meaning of the interrogative sentence to a simple answer set, which on the one hand would ignore the features of the question as a speech act; on the other hand, interrogative sentences are semantically richer than answer sets. For example, in some contexts interrogative sentences with different semantic structures can have the same answer set. Therefore, neither of these two approaches can fully reveal the characteristics of какой interrogative sentences.