Logo DBNL
Voortgang. Jaargang 8 (1987)

Voortgang. Jaargang 8

(1987)– [tijdschrift] Voortgang–rechtenstatus Auteursrechtelijk beschermd

Vorige Volgende

Grammar as a product of text interpretation
Saskia Daalder 1

1. Overview

In this paper I want to shed some light on the intimate relation between grammar and text interpretation in the classical sense of the term. I want to argue that the well-known activities of the explaining and clarifying of text parts can be seen to lead, among other things, to the formulation of a set of regularities which for all intents and purposes may be regarded as the rules of grammar that are in operation. If the argument is sound and if the rather uncomplicated theory of linguistic structure that arises will turn out to be viable, no identification of the rules of grammar will have to be assumed that would be external to practices of text interpretation.

 

The article opens with a consideration of what interpretive statements amount to. The establishment of the relevance and validity of an interpretive statement for an individual text part is seen first to lead to the definition of the elements of that text part (section 2). In section 3 the fact becomes apparent that a relevant interpretive statement very commonly takes account of the text part's elements in ‘phrases’, thus defining those elements into a structure. The notion of ‘subordinate part’ is invoked, leading to a discussion of the so-called inverted constructions with subordinate when-clauses in section 4. Section 5 discusses the ways of validating the structure of a text part as it is implied by the course that a relevant

interpretive statement takes. It is argued that specific text elements from a list of ‘designated’ ones function importantly as indications for interpretive structure. Those phenomena can be taken to constitute the grammatical regularities which are in operation.

2. Text elements and their succession

Interpretation of a text passage by means of a paraphrase and spelled-out details has its place in situations in which someone is thought not to entirely understand the passage, or indeed to understand it incorrectly.2. The function of the explanation is to help overcome the difficulty. The detour of the circumlocutory and explanatory words should restore the text passage to intelligibility.

 

With this in mind, let us consider the make-up of individual interpretation statements concerning a text part. Their abstract form can be represented in the formula ‘text part of description X means Y’. In such an interpretive statement, X amounts to a production-oriented description of some definite stretch of produced speech or script. Y on the other hand is the explanatory expression that should help to improve the comprehension in the situation in which the text part at issue was not fully understood.

 

The first thing to be noted is the irreduceable contextuality of the action of giving an explanation of a text part. It has the definite aim of trying to bring home some specific understanding. This understanding furthermore is usually tied up with a particular perspective3. in which other parts of the text are already being understood or will have to be understood. Within this perspective the interpretive statement for a text part steers towards a particular ‘position’ concerning some topic broached in the text, towards some transient truth which should be taken into account at that specific point. Thus an interpretive statement can be considered as an argument for some particular conclusion. By virtue of the argument status we can use here the terminology of the philosopher and argumentation theorist Naess4. and speak, in relation to interpretive statements, of their relevance or proof-potential for some conclusion.

 

Naess points out that each argument for a conclusion should possess, apart from a measure of relevance, some degree of validity. For each interpretive statement one therefore has to proceed to argue for its validity, i.e. for the truth of the statement viewed in itself. Indeed, how to justify

saying that Y is an interpretation of the text part at issue? Y may be pertinent to some conclusion, but has it anything to do with the text part as identified with the description X?

 

One can try to back up the explanatory expression Y either as a whole, or in segments. In the first case, we have to describe a satisfying number of other interpretive situations where the wording of explanation Y is judged to be of relevance, and where each situation also involves a stretch of talk or writing, as the case may be, whose production may be described as X. That is, we must find evidence of a tradition of understanding in relation to definite features of language production, a tradition to which our particular case can be seen as a continuation.

 

Describing a regularity of relevant text understanding of this kind is of course describing what is traditionally called a language element, such as a word stem, an affix, compound or idiom. We see then that the identification of the existing language elements derives from similarities between cases of text interpretation that have been found relevant. It is on the strength of this fact that the identification can extend into further such cases. The appeal to an element tradition to validate a relevant explanation at the same time also allows for the stretch of production features X to define an element of the text. This comes down to the delimitation within the text of an instance of a word stem, affix, compound, or idiom.

 

In the case of a validation in segments, we are obliged to give such a demonstration separately for each ‘explanatory part’ to be distinguished within the comprehensive explanation Y. But now we have to deal also with a chosen ordering of those individual explanatory parts within the comprehensive Y. This internal ordering may be of the utmost importance for the proof-potential of the explanation, because in fact it often involves a different wording of the individual explanatory parts.

 

As an example we may consider Bolinger's (1952: 1120) explanation of the Spanish sentences Juan canta and Canta Juan. As for the first, a relevant explanation starts with information about someone called Juan, and goes on to specify that he is a singer or is at least involved in singing rather than in something else. In the other case, the issue is explained as an action of singing that anybody might be doing, such as in a concert of several singers, specified subsequently by the information that it now is Juan's turn to sing rather than Maria's or Pedro's.

Thus the explanation for Juan canta deals with an interest in the doings of an individual person and defines the form of one of his activities, while in case of Canta Juan the explanation concerns the experience of a particular activity and how to distinguish and anticipate the performances of the individuals involved. The fact that these explanations do not amount to the same thing testifies to the absence of a fundamental hierarchy in the dealings with the two aspects involved, in spite of the fact that they are strongly internally related. It is apparently not the case that the traditions of dealing with persons essentially would be derived from traditions in dealing with activities; nor does the opposite dependence hold. The ordering of the explanations of Juan and of canta is just something of one individual situation in which they both happen to participate. That is, it is not governed by a hierarchy or independently existing ‘structure’, but rather teaches the best succession of going about managing the different aspects within that particular situation. It can generally be said that the ordering within a comprehensive explanatory expression represents that particular succession of dealing with the interrelated facets of the situation which leads to the conclusion deemed useful for subsequent comprehension.

 

Returning to Bolinger's sentences, we can say that the pertinent explanatory ordering in each case defines an ordering of the corresponding text elements. It defines a ‘path’ of ‘interpretive steps’ through the text. The following diagrams indicate these paths by means of consecutive numbers written under each interpreted element.

(1) Juan canta
  1 2
 
(2) Canta Juan
  1 2

These examples already suggest which language production aspect is principally used for the validation of the explanatory ordering of elements: their linear or temporal sequence. The more we follow, in our explanation, the sequence in which the elements of the text part were produced, the more valid becomes the assertion that the relevant explanation we are giving concerns the text part in question. And of course, to the extent that a relevant interpretive ordering of elements is validated by their production sequence, to that extent the linear order of the utterance elements can also be said to directly indicate the order in which, for best results, to tackle the interrelated aspects of the situation represented by

each element. To that extent then Bolinger's (1952) principle of ‘linear modification’ is vindicated.

3. Structure

However, the interpretive construal cannot be limited to adding explanatory expressions one by one for each text element. Not surprisingly, it also needs the formation of ‘phrases’. To obtain relevant interpretations we very commonly have to take interpretive steps which bear on several text elements together. This represents a structured engagement with the individual facets of a situation that are indicated by the text elements. The required interpretive technique can be conceived in a very uncomplicated fashion, as follows. We first construct in individual element steps the explanation of a phrase composed of several individual elements, and then simply use the completed expression as a further specification of another explanatory expression that we have given independently.5.

 

For illustration, let us first look at the very simple example (3). Here we do not want to say that the explanation of in would be specified by the interpretation of the by itself, the outcome of which would only then be specified further by an explanation of city. That procedure would ignore the extent to which an explanation of the city, quite apart from an interpretation of in, could clarify many aspects of the dealing and engaging with this thing that the city is. Hence we must first compound the interpretation of the phrase the city, as is represented by the first line of diagram (3). The outcome of the last step of this line (step 2) should then be used to further specify the explanation of in, which itself constitutes the first step of a second line of interpretive steps. To indicate that the outcome of the preceding line is taken up with the second step of this new line, this step is represented directly beneath the preceding line's last step.

(3) in the city
  1 2
  1   2

Of course a phrase may be part of a text part that eventually is itself explained as a phrase within the next larger sequence of elements. The more complex diagram of example (4) represents such a case. An interpretive phrase here corresponds to some further step of the interpretive path

associated with a containing text part that itself again corresponds to a single further step in the interpretation of a comprehensive text part.

(4) Be careful when crossing streets
  1 2
  1   2
  1 2   3

First the interpretation of the phrase crossing streets is constructed; this is represented by the first line in the diagram. The compounded explanatory expression is then used as a further specification of the interpretation of when ‘indeterminacy as to circumstances’, as is represented by writing step 2 of the second line directly under the last step of the preceding line. The explanation resulting from this is something like ‘indeterminate occasions having to do specifically with crossing streets’. This expression is used subsequently to specify the advice to take care (formulated independently - steps 1 and 2 of the third line) as bearing on such occasions. This last move is represented by step 3 of the third line, written under the last step of the preceding line in order to indicate that it makes use of its outcome.

 

It may be noted that with this conception of structure the basic temporal or linear progression of the interpretation can be maintained all along - on each diagram line there is normally a regular succession of consecutive numbers. Thereby the interpretive path can derive as much validation as is possible from the linear aspect of language production. The structuring is thus a modulation on the linear movement rather than a violation of it.

 

‘Phrases’ as defined in the above exposition are characterised by internal independence coupled with external subservience. This is a strong reminder of the situation commonly considered as typical of subordinate constructions. There is a logic then in using some of the subordination terminology and in investigating the question as to whether the approach to structure described here could solve problems relating to phenomena in the field of subordination. Let us first define the term subordinate part in the following way.

 

(5) Definition subordinate part.

When an explanatory expression which bears on several text elements is used as a further specification of another explanatory expression,
the text part at issue is called subordinate.

For the examples discussed so far the definition has the following consequences. In diagram (3) the explanation of the city is seen to function as a further specification of another explanatory expression, and the city is thus a subordinate part. Diagram (4) shows both crossing streets and the encompassing when crossing streets to have the required property of an internal interpretive independence that does not stand in the way of an external subservience of the kind mentioned. The explanation of the whole of crossing streets serves as a supplement to the explanation of when, while the interpretation of the complete when crossing streets adds to that of be careful. Thus crossing streets as well as when crossing streets count as subordinate parts.

4. Inverse when-constructions

This latter fact may lead us to an exploration of a problem within the field of subordination that also concerns constructions with when.

 

The phenomenon of subordinate clauses that seem to take a kind of precedence over their main clauses is well-known. They assume for themselves several of the properties typical of main clauses, in particular those having to do with emphasis and assertion. At the same time the role of the real main clause in such a case seems to be reduced to an indication of presuppositions or circumstances of the fact related in the subordinate clause. Among those problematic constructions are the so-called ‘inverted’ constructions6. with a when-clause such as exemplified in (6).

 

(6) He had not walked three paces, when he turned angrily around

 

Let us try to determine the structure of this sentence, in the sense of the ‘path’ which a relevant interpretation follows relative to the elements of the sentence. The path associated with the when-clause of (6) in fact can be seen in one crucial respect to be different from the one used in example (4).

 

We can first say that the interpretation of the text part following when in example (6) does not function as a further specification, independent in itself, of the indeterminacy as to circumstances that is the interpretation of when. Rather, this indeterminacy must be explained by that which has

been interpreted in relation to the preceding text part: his decision to walk off and in particular its execution in the form of the completion of individual paces. With when they are referred to as circumstances so precarious as to leave room for several different sequels. Now with the subsequent interpretation of the elements following when this less than well-defined situation is seen in fact to develop into the event of his turning round angrily. That is to say, the elements following when are from the outset interpreted under the terms of the operation of the insecure situation. Therefore, in contrast to crossing streets in (4), the text part he turned angrily round does not have the interpretive independence required by the definition of subordinate parts. Cf. diagram (7).

(7) He had not walked three paces, when he turned angrily round
  1 2 3 4 5

However, the fact of the development of this particular event from a situation that could also have developed differently is used subsequently to further specify what is interpreted from the initial clause. It is an addition to his not having completed more than a few steps in walking off, typically presenting a reversal of implications. The complete when-clause (i.e., inclusive of when) therefore has to be indeed considered as a subordinate part, whose interpretation is formed independently and then further specifies another explanatory expression. We indicate this by continuing the interpretive path associated with the initial clause with a further step: step 3 on the second line in interpretive diagram (8) of the full sentence, written under the last step of the interpretation of the when-clause.7.

(8) He had not walked three paces, when he turned angrily round
  1 2 3 4// 1 2 3 4 5
  1   2   3

We see then that, in the approach presented here, the ‘standard’ when-clause of (4) and the ‘inverted’ one of (8) differ as to the internal structure of the when-clause. This clause comprises an internal subordinate part in the first case but not in the second. Meanwhile, the subordinacy of the complete when-clause can be maintained for both cases. To bring out the distinct character of the ‘inverse’ constructions with a when-clause like (8), it is thus not at all necessary to resort to a full-scale reversal of the relation between their clauses such as would be the effect of

attributing a subordinate status rather to their initial clause (cf. note 6 above).

 

Thus it seems that the idea of establishing the structure of a text part as a corollary to its explanation, along lines such as those sketched in this paper, holds some promises for solving long-standing structure problems.

5. Grammar

For all texts or utterances, the reason for following one interpretive path rather than another lies in the eventual relevance of the resulting interpretive statement for some further conclusion fitting in a broader view of the text. This is not to say that the structure which is thereby attributed to the text would be something subjective. To quote Potts's (1977: 94) Wittgensteinian phrase: ‘Structure is not subjective, but relative - relative to a purpose’. The purposes of one's theoretical activity naturally make themselves felt.

 

But the claim that one is interpreting a particular piece of text leads to the obligation to indicate valid connections between aspects of the interpretive structure and selected textual production features. We have seen already in section 2 that the linear sequence of the text elements provides the basis for validating the chosen progression of interpretive steps. However, this *only holds within a given interpretive phrase. The structural aspects of the interpretive path clearly cannot be backed up in this manner.

 

What kind of validation can in general be found for the actual phrasing of interpretive paths? It will suffice to answer this question for the deviations from a strictly linear interpretive processing of a text's elements. As it turns out, those deviations can suitably be identified with relevant interpretive moves of treating an individual text part as a subordinate one in the sense explained in section 3. The question therefore is: which specific features of language production can be found regularly to concur with what is a relevantly subordinate part?

 

Typically, such validation of structure8. may involve features of prosody or punctuation, present at crucial positions within the text part. Specific graphic devices of typeface or page layout may play a similar role, for instance in poetry or on posters and the like. Structural regularities can

be formulated that each describe a tradition of interpreting a text part as a subordinate one in connection with a specific position, relative to the text part, of certain such production features that are not themselves associated with explanatory parts.

 

But validation for interpretive structure can just as well be furnished by production aspects that also relate to individual parts of the structured explanation. It can be provided by the presence of specific individual text elements at well-defined structural positions within the text part.

 

Let us illustrate the latter fact with cases already discussed in the previous sections. In example (3), in the city, the text element the introduces a text part which is relevantly interpreted as a subordinate one (the city). This is because of the fact that its elements are interpretively ordered among each other but not with respect to the element in, which is outside of the subordinate part. It needs no defense to say that such an interpretive position is something which is very often found in connection with instances of the. A subordinate interpretation of a text part starting with an element the regularly leads to good interpretive results.

 

Similarly, in (4) as well as in (8) (repeated here for convenience), instances of the specific element when open up what we have seen are relevantly interpreted as subordinate parts, when crossing streets and when he turned angrily round respectively. Again we can say that these cases continue a tradition, in this case a regularity of interpretive structure in connection with text elements when.

(4) Be careful when crossing streets
  1 2
  1   2
  1 2   3

(8) He had not walked three paces, when he turned angrily round
  1 2 3 4// 1 2 3 4 5
  1   2   3

It seems in fact that regularities of just this format could also be formulated in relation to other elements ‘designated’ in grammars as somehow regularly participating in the structure of utterances and text parts. All cases

indeed seem to involve some specific text element on the one hand and some position of this element relative to a containing text part on the other. Flectional elements, which are usually not written as separate words but rather appear as modifications or ‘endings’ of other elements, satisfy this description just as easily as text elements of independent occurrence.

 

We can appropriately use the term rules of grammar for all regularities that describe traditions of interpreting a text part as a subordinate one in connection with some well-defined position, relative to the text part, of such a ‘designated’ text element. Although much exploration still has to be done, it does not seem impossible that all phenomena traditionally considered as belonging to grammar will turn out to fit in this scheme.

 

Grammar, we could then say, concerns the coordination of ‘designated’ text elements with relevant interpretive structure. And as a natural consequence of the fact, explained in section 2, that certain similarities between adequate cases of text interpretation themselves lead to the identification of the language elements, this would mean that the rules of grammar are also inherent in text interpretation. The identification of the rules of grammar which are in force would likewise derive from cases of interpretation found to be adequate within some broader perspective. The rules of grammar would be constituted by similarities, between those cases, in details of the coordination of instances of a designated text element on the one hand and relevant interpretive structure on the other.

 

There would be no need then to think that grammatical rules would determine the structural aspects of utterances prior to or apart from interpretations of those utterances which could be judged relevant for some conclusion. The statement of grammatical rules would have to be viewed rather as resulting from collections of many particular cases of adequate text interpretation. In its correlation with the positioning of specific text items, the relevant interpretive structuring of such interpretive statements would form the basis for the formulation of traditions of relevant grammatical structure. Grammar could be viewed as a product of text interpretation.

References

Becker 1984:
A.L. Becker: ‘The linguistics of particularity: interpreting superordination in a Javanese text’, Berkeley Linguistics Society 10, 425-436.
Bolinger 1952:
D. Bolinger: ‘Linear modification’, Publications of the Modern Language Association of America 67, 1117-1144.
Bolinger 1977:
D. Bolinger: ‘Another glance at main clause phenomena’, Language 53, 511-519.
Daalder 1986:
S. Daalder: ‘Element en structuur: Nederlands toch en doch’, in: Syntaxis en lexicon. Veertien artikelen bij gelegenheid van het emeritaat van Albert Sassen, Dordrecht and Providence: Foris, 132-156.
Green 1976:
G.M. Green: ‘Main clause phenomena in subordinate clauses’, Language 52, 382-397.
Naess 1966:
A. Naess: Communication and argument. Elements of applied semantics, Oslo and London: Universitetsforlaget and Allen and Unwin.
Naess 1978:
A. Naess: Elementaire argumentatieleer, with an introduction to the philosophy of Naess by E.M. Barth, Baarn: Ambo. Basisboeken.
Pos 1923:
H.J. Pos: Kritische Studien über philologische Methode, Heidelberg: Carl Winter. Beiträge zur Philosophie 10.
Pos 1933:
H.J. Pos: ‘Ueber den Aufbau der grammatischen Interpretation’, Blätter für deutsche Philosophie 6, 295-314.
Potts 1977:
T.C. Potts: ‘The place of structure in communication’, in: G. Vesey
(ed.), Communication and understanding, Sussex and New Jersey: Harvester and The Humanities (Royal Institute of Philosophy Lectures 10), 91-115.
Quirk 1972:
R. Quirk et al.: A grammar of contemporary English, London: Longman.
Quirk 1985:
R. Quirk et al.: A comprehensive grammar of the English language, London and New York: Longman.
Uhlenbeck 1962:
E.M. Uhlenbeck: ‘De beginselen van het syntactisch onderzoek’, in: Taalonderzoek in onze tijd, Den Haag: Servire, 18-37.

Vorige Volgende

Over dit hoofdstuk/artikel

auteurs


taalkunde