Decidability Problems in the Humanities

by Jos Kunst (1986)

Originally published in Liber amicorum J.L. Broeckx; [ed.: Ferdinand J. de Hen, Anne-Marie Riessauw, Herman Sabbe]. Gent: [S.n.], 1986, p. 195-206.
Slightly modified version: On methods in analysis and empirical research. In: Performance theory; Henri Schoenmakers (ed.). (Advances in reception and audience research; 1). Special issue of: Tijdschrift voor theaterwetenschap: 16-17 (1986) 7-17.
According to the author, the version in Performance theory contains errors because of unwarranted editorial corrections.

0.0 Ideologies

Let us take the term 'ideology' in the broadest of its current senses, and analyze it as the set of all strategies people have (possibly in common with all or most members of some community they belong to) in order to come by ideas, to find statements to make, acts to perform, thoughts to entertain, etc. It is then clear that, like all social endeavours, the scientific enterprise may be said to be driven by the (explicitly or implicitly shared) ideology of the members of the community involved in and defined by it. Obviously, this goes for natural sciences, social sciences and humanities alike, whatever the degree and kind of empirical control they may have developed. At the core of ideology so described, perhaps even as its main ingredients, we then have concepts, provided we analyze these also broadly, viz., as recognitional capacities, without any such traditional restrictions as high degree of abstraction, verbal nature, etc. In this way it is clear why there can be no easy exchange of a given ideology for another, such as is at least thinkable for any explicit opinion, or set of such. After all, concepts are necessary ingredients of even mutually exclusive beliefs, and therefore less easily uprooted.

Why do we go into this ? Because in scientific communities, even before any rational discussion, their shared ideology (co)determines what any rational discussion can be about, and, in the cases where there is a choice, the direction this choice actually will tend to take. All this is well known, and well documented as well (cf. Kuhn 1970 and the evaluation of the Kuhn-Popper discussion in Stegmüller 1973, but also Habermas and the French structuralists).

0.1 Conventions

Whether this 'limited rationality' actually boils down to simple irrationality depends on how one decides to use the words involved. One thing is certain: it is that within that part of the humanities which is involved in the interpretation of texts and the analysis and evaluation of art the only traditional check on what one may respectably assert consists in the question whether or not, within the scientific (or, before the scientific age, the literate) community a convention exists or may be reached with respect to it; most of the time, this will be so only if it can be accomodated within existing conventional frame-works, and, preferably, confirm and strengthen them.

What must interest us here is the question whether conventions can be said to be rational, and if so, in what sense. According to the famous analysis by David Lewis (cf. his 1969) a convention is the solution to a coordination problem. A coordination problem is a situation in which two or more people are actively engaged and which will be beneficial to each of them if and only if it will be beneficial to all (or almost all) of them. Under some description of their behaviour, all of them will have to do 'the same thing'. There is a problem if there is more than one behaviour that, if only adopted by all (or almost all), would do the trick. The problem is solved if there exist, to the knowledge of the people involved, precedents for it, earlier situations sufficiently like it, and known by all to be known to all, in which some known behaviour has helped them out. They will then choose that same behaviour again.

0.2 The conventionality of traditional humanities

Globally speaking, one can say that any participant in a given culture has a conversational need of a certain kind of self-confidence: that in making assertions about cultural deeds or artefacts he will not cover himself with ridicule or otherwise prove himself to be a cultural outsider. In certain respects he wants to say precisely the same thing as all (or almost all) the others do, e.g., produce the same kind of variations on a limited set of fixed themes. Literary criticism, and art criticism in general, have, since time immemorial, been trying to provide the right kind of precedents. They have, since not so very long ago, produced an academic offspring, in which scientific authority and, sometimes, taste arbitership combined to produce the right kind of 'salience' needed for opinions to become suitable convention-governing precedents.

(One important caveat: conventionality does not imply anything like absolute arbitrariness. As is shown in Rollin 1976, the conventional and the 'natural' are not separated by any real gap: on the contrary, there is a continuous gamut of possibilities in which the purely conventional and the purely natural only occur in the extreme (and idealized) positions: everywhere else, i.e. in normal space, convention operates within varying margins which are themselves not conventional).

1.0 Interpretation

We may be approaching something of a quagmire. Obviously, literature itself is a conventional affair. Moreover, literary criticism is itself part of literature, its arbitrariness being hemmed in by the fact that it must not only fit in with, but also be in some way recognizably about, the rest of literature, whereas the rest of literature may be just about anything. Now, in the (natural) sciences the 'scientific enterprise' may be largely convention-governed, but its object 'in itself' clearly is not, and science operates in such a way as to permit its object to effectively 'feed back' on it: that is the essence of its empirical nature. In the human sciences, on the other hand, some of the most interesting subjects for research are formed by conventional structures, e.g., those of art and literature. Are we in trouble if we are, as we seem to be, approaching convention through convention, e.g., the conventions of literature through those of literary criticism, which is itself part of literature?

1.1.0 Literature on literature

In the first place: there is, I think, no 'mixed levels' problem here that would give rise to any logical paradox. Grammars of English are, most of the time, written in English and none the worse for it (see, on the ways these paradoxes are avoided in 'natural' communication, Fogelin 1976).

But, in the second place: it may not be easy to see why the academic offshoot of criticism should enjoy a social status, such as expressed, e.g., in the amount of money per person invested in them by the community, that is as markedly superior to that of the undomesticated forms of literature as it actually happens to be. In other words, on what special grounds do the academic professors of literature earn their salaries?

Obviously, their degree of literacy is not higher than that of their colleagues outside academia. If they are historians, they may know more about specific periods, but then it is to that historical competence that they owe their status and their money. Neither are they more productive. But what they do write is assigned a special, viz., scientific, status, which seems to imply something like a higher degree of certainty. The question remains: on what grounds ? They have, in general, ceased to decide on matters of taste, but they still do decide (or try to) on matters of interpretation.

1.1.1 Hermeneutics

In questions of theology it does not strike me as strange (not stranger, anyway, than theology itself) that certain people are singled out, and permit themselves to be singled out, to bear the burden of authority. Some authoritarian way of fixing the coordinated behaviour of a religious community is needed; scriptures are there, and they are quite ambiguous, to say the least. Honours and high pay are simply the community's way of concentrating authority in a small group of its members. The methods employed by these authorities are both, I think, as reasonable as possible and in no way special. What they amount to is essentially informal imaginative thinking about either, in contemporary perspective, the ways a text can be plausibly understood, or, in a historical perspective, the way, or ways, a text may have been understood by its author, or both. I contend that this is, in essence, what any reader does with any text.

1.1.2 Literary theories

The central problem of hermeneutics is, I think, the heuristic (and ideological, cf. 0.0 above) one of coming by ideas, or (informal) theories. Now clearly, these ideas, or theories, will often be mutually incompatible, and the degree of tolerance people have for entertaining mutually incompatible theories is both real and limited. It could even be argued that the 'uncontrolled', i.e. only socially limited, production of mutually incompatible theories belongs more properly to the province of art than to the province of science (that is why hermeneutics is often taken to belong to art). Once there is something like an explicit and controllable method involved, there may be a decision procedure in the offing: a result may, and on the other hand also may not, be warranted by the method.

Several kinds of semiotics do claim to have such methods (cf., e.g., for music, Nattiez 1975). If they are right, they have one quality traditionally associated with proper scientific endeavour, viz., inspectability. Inspectability is, I take it, cast in that role because of the factor of intersubjective control inherent in it: even where the nonscientific world has to take on trust the results of science, it counts on the scientific community as a whole to exercise some control on individual researchers, in such a way that these will be effectively prevented from publicly saying things because they just happen to feel like saying them. Surely, there is no such thing as 'disinterested' research; but, the argument runs, if it cannot be determined by the interests of the social system as a whole, let it at least be guided by those of the scientific community rather than by the personal quirks of single individuals. Why is this important? In what ways do interests and theories relate?

Mooij 1979 provides us with a lucid analysis of the philosophical status of most current theorizing within the field of literature. In accordance with many of the more recent philosophers of science he makes a distinction between the explanatory and the exploratory functions of theories. Theories are not only, and theories of literature are perhaps not even in the first place, characterizable by their propositional content, but also, and in the case of literary theories even chiefly so, by their conceptual structure, by the fact that they provide you with new 'ways of seeing', new ways of 'informing the literary eye', new ways of thinking about literature. Obviously, this is what literary criticism and, more generally, interesting comments on literature have always been trying to do. Since it is not plausible that professors of literature are more inventive than their non-academic predecessors and colleagues, their special quality must lie rather in ways they have, 'relevant' ways, of keeping in check the fertility of their imaginations. Toward the end of his essay Mooij takes up the issue of how literary theories may come to fail. He mentions three possibilities (cf. p. 134):

  1. by stagnation: a theory then does not yield enough interesting new problems;
  2. by infirmity: it then yields too many problems which it is unable to treat; and
  3. by incoherency: the treatment of the problems it yields is, in that case, not adapted to the way they were put in the first place.

As I understand them, these are also the ways in which, within science, whole paradigms, rather than individual theories, may get into trouble; anyway, the decision seems here to be between fruitfulness and barrenness rather than between rightness and wrongness.

The way a theory may survive is precisely by avoiding 1., 2. and 3., i.e. 1. by continually raising 'interesting' problems; 2. by coming up with solutions for them; and 3. by coherently enriching itself with these solutions. A theory (which is then, obviously, equated with a conceptual scheme, a heuristical device) must grow in order to survive.

My intuition tells me that this description does indeed have elegance and convincingness, but also that it applies uncannily, and uncomfortably, to literary fashion itself.

This, we want to argue, is not a satisfactory state of affairs. When semioticians, by suddenly styling themselves as Peircean instead of Saussurian without being able to specify the empirical consequences of such a step, when they try to launch in this way something like a new fashion in semiotics, just by exchanging one conventional authority (precedent) for another, I feel that what is at issue here is perhaps nothing more than an extremely elitist literary revolution, and this makes me feel uncomfortable indeed. All this, it seems to me, is bound up with what Mooij calls the 'plasticity' of literary (or musical, or theatrical, or what-have-you) works. Obviously, there is no limit to the number of interpretations that can be associated with them. But, at the same time, I think there is a limit to the interpretations a given culture is in fact able to associate with them at any given moment. For there interpretations will have to be coherent with the whole system of conventions that make up the culture, and in, e.g., reading a text, people might well be trying to accomodate within their understanding of it as much of their culture and knowledge-at-large as they can. Therefore I think that this 'plasticity' of the art object might well be nothing more than an illusion based on a kind of 'telescoping' of moments in history, or possible worlds – a thought no one would seriously entertain if he were not, in some way, conceiving art works as 'independent' and somehow self-sufficient entities. All this, I think, warrants a closer look.

2.0 The 'locus' of art works

There has been, as my readers well know, rather much discussion about what a work of art actually is. This discussion might, in the long run, prove quite unfruitful if there were no, or almost no, or only an extremely trivial, constant content associated with any given art object.

I want to side-step this problem (I think the answer will vary from one art object to another) and, instead, ask a different, and empiricism-oriented, question, viz., where are art works?

If anyone puts forth a statement about one or more of them, where can I go and see wether he is right or wrong?

2.1.0 Memory

The answer we give here obviously wants to propose for future use a scientific convention, implying a way of looking at art works, as do the others. But I think it has, in addition, at least the advantage of guaranteeing us relevance and empirical content. Art works, just like signs generally, we argue, are localized in the memories of their users.

2.1.1 Internal and external memory-spaces

A first important caveat: this localization of signs and art works, perhaps surprisingly, does not yield in any way a poorer, in the sense of smaller, collection of them than the traditional one, for human memory itself has two kinds of loci, viz., inside and outside the skull.

Humans do use not only internal, but also external memory-spaces. Some people tie knots in their handkerchiefs in order to be reminded of something. Other have strings of numbers written on crumpled pieces of paper in their pockets. Many use appointment books, and personal lists of phone numbers. These are still relatively private external memories. But phone directories, public libraries, and other shared and public artefacts such as theatres and towns are very much less private already. Still, the fact remains that within the internal memory-space there must be something like a 'program' specifying the access to the external one. A child, in general, thanks to its internal programming, cannot help picking up the natural language of its surroundings, and this, in turn, provides him with a program ensuring its access to further parts of the public memory-space, such as libraries, etc. (Also, extensions of the 'working memory' information processing theorists talk about, are used by humans, e.g., in doing long hand division on paper: an external working memory).

It is, obviously, within these 'external' memory-spaces that we find things like signs, art objects, etc., as they are more traditionally thought of; equally obviously, by taking such a stand we have in no way impoverished our supply of them. What we did change is our conception of them.

If their locus is the human memory-space, no physical description of them will be adequate. Instead, we will need a 'psychological' description in some extended sense of the word. Signs and art objects will be mental objects. But they may also be public objects, because the mental space they are in may be an external and public mental space. And here comes in our second important caveat: this external and public mental space will have to be defined by the reference to some existing social-cultural group. Only the members of that group will have the proper access to their shared external memory-space.

This is theory. It supposes that the use of some external memoryspace may, within a given sociocultural group, be sufficiently codified and coordinated to permit their description as actually shared memory-spaces. But it is both reasonable and testable. Mooij's 'plasticity' of the literary work (see his p. 131) is seen in its light to be nothing else than the plasticity of the observer, who may, all the more so as he works in a more isolated position, manipulate his own memory structures in such a way that a given art work comes to present many divergent aspects. He may 'go too far' in this: i.e. incur the risk of losing contact with his group, of leaving the shared memory-space. Therefore he may try to persuade the others to take over his viewpoint, thereby reestablishing contact. He then, obviously, functions as a literary artist writing on literature; he is a 'researcher' only in the extended sense in which certain avant-garde artists have claimed to be such.

At least an important part of academic endeavour with respect to art should concentrate on its functioning as a means of communication within the social group that uses it, and try to ascertain which concepts are centrally, and which more marginally, or possibly not at all part of the group's shared memory-structures and hence provide, or fail to provide, a reliable basis for communication.

2.2.0 Art theory

This does not mean that reliable communication does nothing else than use existing conceptual structures. In Kunst 1978 we have argued at some length that art typically involves an 'interestingness' claim, and therefore initiates what we have there called 'unlearning/learning processes', a deconditioning-cum-reconditioning if you wish, incorporating a strategy of attracting attention to an in some way special behaviour: this behaviour is such that a. it constitutes a transgression of some conventional law; that is how it can attract attention (and provoke affect at the same time); and b. it constitutes in itself a precedent for a new conventional perspective wherein a new law-likeness, and therefore a new understandability of subsequent behaviour is made possible.

2.2.1 Art as we know it

Now many behaviour that is perceived as intended to be perceived as described under a. and b. is called *art* (between asterisks) in Kunst 1978; it might perhaps be described more aptly as art in statu nascendi-, art 'being learned' perceptually; that is why it is of crucial importance to any 'genetically' oriented theory.

On the other hand, I do not think that this 'genetic' twist to the theory makes it otherwise uninteresting. It could be argued that well known and functioning art works, precisely in as far as they continue to be interesting, confront their users again and again with unlearning/learning processes that seem well-known not because they could be recalled, but because they are recognized in passing and therefore gone through again and again; moreover, 'old' art works may well, in new settings, initiate unlearning/learning processes that actually are novel: their users are then confronted with Mooij's 'plasticity of the art work': i.e. their own plasticity.

More generally, one might ask: how are art works remembered? What is one thinking of when one thinks of art works? – and argue for the idea that it is for their effects, for what they did to you, that art works are remembered; that, in turn, would mean that they are stored in memory under a pragmatical description – how they are remembered is then determined by the way you got acquainted with them in the first place: by the unlearning/learning processes associated with them.

2.2.2 Art research

Kunst 1978 provides a partial formalization of this process of getting acquainted with (aspects of) art objects or even events which is particularly useful for the purposes of the empirical research advocated in the last paragraph of 2.1.1. This is so because we find there combined within one and the same frame descriptions of three different domains, viz., 1. a description of (an aspect of) an art object or event itself as it is perceived, 2. two descriptions of the conventional background used in 'placing' the art object or event: part of the domain of social-cultural facts, and 3. four descriptions, forming two pairs, of the way the art object or event is, under the given description, 'placed' against the two given conventional backgrounds: this is the domain of the individual information processing. (There are two here for each conventional background because what interests us here is the difference between not having the information and having it; there are two conventional backgrounds because of the double nature of the unlearning/learning process.) Now what is important to us here is that these descriptions, seven in all, are interconnected, on theoretical grounds, a. by the rules of a modal logical calculus, viz., the Lewis-Langford system S7, and b. by a number of additional assumptions laid down in the properties of the modal frame in question and in the relations between the seven 'possible worlds' that make it up. These theoretical riches make for empirical usefulness: they may suggest, in circumstances which are not fully known, what to look for, e.g., in social-cultural backgrounds; they bring up questions to be asked of the art users involved; finally, they provide descriptions of the art activity 'itself' (descriptions of an 'interpretational' nature) which correspond to social-cultural and perceptual contexts. This very general (i.e. only *art*-specific) set of theoretical assumptions is, in Kunst 1978, laid down in a structure called the Bivalence Function, which maps ordered triples of propositions describing *art* behaviour onto the seven-worlds frames detailed above. The Bivalence Function schema itself uses variables, in the sense of place-holders: any given context of research may determine what will be substituted for them, and the result is an in principle testable theory. Different substitution instances (called 'BivFs') may be, in their turn, interconnected in specific ways to form 'pragmatical networks', which make it possible to produce theories in which not only, say, verbal, musical, gestural and stage effects, but also lower level and higher level ('ideological') perceptions may be integrated.

3.0 Conclusions

  1. On the theory formation side of research into art objects or events anything goes, be it hermeneutics, Verstehen, semiotics of any brand, or whatever.
    However, if scientific status implies, as I think it does, something like a rationally arguable higher degree of certainty, there had better be something on the theory testing side as well.
  2. It is advisable, in the heuristics of theory formation, to strive for coherence with existing theories, the more so as they have fared better in testing situations: this not only makes for better initial plausibility, it also helps us in the often unfamiliar heuristics of test design.
  3. The proper testing ground for art theories are the existing art users and the social-cultural world they live in: this makes for intentional terms in the theory.
  4. This intentionality is captured, in Kunst 1978, not only by the use of intensional logic, but also by the specific properties of the modal frame described in the Bivalence Function. Moreover, this intentionality is conceived there as a shared and therefore social-cultural fact, and hence open to empirical research.


Fogelin, R.J. Comment on Hintikka's paper. In: Koerner, S. (ed.), Philosophy of Logic, Oxford, 1976, pp. 233-243.

Kuhn, T.S. The Structure of Scientific Revolutions. Chicago, 1970.

Kunst, J. Making Sense in Music, an Enquiry into the Formal Pragmatics of Art. Ghent, 1978.

Lewis, D. Convention, a Philosophical Study. Cambridge Mass, 1969.

Mooij, J.J.A. The Nature and Functioning of Literary Theories. In: Poetics Today: 1 (1979), pp. 111-35.

Nattiez, J.-J. Fondements d'une sémiologie de la musique. Paris, 1975.

Rollin, B.E. Natural and Conventional Meaning. The Hague, 1976.

Stegmüller, W. Theoriedynamik. Berlin/New York, 1973.