[61a] Philip Pilkington, ‘Some Darwinian Conditions of the Art & Language Indexes’, Art-Language, New Series No. 2, June, 1997, pp. 3-11.

Some Darwinian Conditions of the Art & Language Indexes

Philip Pilkington was involved in the Art & Language Indexes and now works in higher education. He is also Chairperson of the Transport and General Workers Union (TGWU) at Coventry University, Secretary of the Joint Union Committee, and adviser on higher education to the TGWU.

This paper is not a remembrance of time past. It is a somewhat diagrammatic and canonical explanation of the genera Art & Language (AL) Indexes. It does not provide a chronological survey or suggest an evolutionary unfolding of the AL Indexes between 1972 and 1976. It does unpack some of the historical preconditions and the subsequent commitments which the AL indexing projects represent. The generalisations found here result from an aim on my part to describe a genealogical ascent. If the paper appears somewhat abstract or detached from the struggles within the historical practice, it should not be presumed a friction-free vanity-press idealisation, but merely discursive.

The antecedent conditions of the Indexes were manifold: pedagogical relations, both formal and informal, collapsed into non-hierarchical conversations which internally and externally constituted critical discourse; resources of description and explanation were developed, mutated and borrowed from a large and eclectic range of philosophical sources which constituted a critical internal (sic) function; there were some who entertained an embarrassingly corporate ideation of AL practice which was not discursively reconcilable with the actual conversations; the cultural superstructure requisitioned and appropriated pro forma AL practice into the contemporaneous art practices dubbed Conceptual Art and AL was in no position to resist. These conditions did not stand in discrete relations to one another (logically or temporally) but were reactive and dynamic.

It is useful here to consider the claims made by and made for the neighbouring artistic practices of ‘Idea Art’ in order to provide some contrast to the Indexes. It should be noted, however, that such claims were a minor theme in the pedagogic conversations at Coventry Art School in the early 1970s which were concerned with the cognitive and political questions of practice. This last remark should not be taken as an attempt to certify some sort of methodological sophistication or rigour or distancing on AL’s part. (It will be helpful in this connection to keep in mind the possibility that the Indexes were realist and representational.) There was in the conversations of AL a curiosity – a practical curiosity – which seemed to be of a more complex order than could be discovered in the actual corpus of work known as Conceptual Art. This curiosity was concerned with the question ‘how do you deal with’ (reconstruct, use, cope with, etc.) contemporary claims made about ideas (as art)? (1) This questioning was not pursued in order to legitimise or entrench an apparent terminus of modernist reductivism and so be terribly avant-garde. It was a ‘scholastic’ interest not in de facto legitimacy but in the problem of what might be counted as thought in, and constitutively for, AL practice. The internal dialogical nature of AL practice meant that there could not be a proprietorial and thus reificatory hold on ideas, thoughts, etc. The external problem grew (apart from a salesman’s failure to discriminate between thoughts, ideas, propositions, etc) from a general assumption that AL and Conceptual Art shared implicit historical practices. The naïfs from the neighbouring practices offer a contingent and fortuitous contrast to AL’s discursive and methodological roots. The historical problematic of ‘ideas’ (etc.) for the vanguard of late Modernism was imprisoned in the late nineteenth century. The naïfs exhibited (literally) the by-blows of emotivism and the doubtful authenticity (aesthetic and otherwise) of what’s inside you (and your head) – and notwithstanding the look of the production. The opportunism of exhibiting (literally) the ‘mental life’ of artists was a reification consistent with panpsychicalism. This condition was neither understood nor probably understandable by its perpetrators, who were playing with Meinongian fire. They were, thereby, unwittingly exemplifying the problem of the intensional – of Meinongian-Brentanoesque ‘objects of thought’. They were not concerned with the genuine problems of ontology. They were unaware that intensionality involves ontological problems of existence and the quantificational problem of identification, functionality and the meanings of thoughts, ideas, etc.

A possibility that was inescapable in twentieth-century (philosophical and scientific) thought was that you either had Meinongian objects of the brain/mind without any clear view of how those objects interrelate (they just are), or you had thoughts as complex functions with (classically) sense and reference as the constitutive meaning (pace Gottlob Frege). This was (and is) not a rationalisation or critique of aberrant art and art talk, putting people in their place; the Fregean picture of thought as a functional reference has a general and powerful application to language and meaning, and the historical discourses used by AL provided an understanding of our own and others’ linguistic, conceptual and cognitive practices. This was an intellectual (or just intelligent) step function for AL from the realisation that meaning is a product of reference (to things individuated by a thought), or that the existential condition ‘x has thought that__’, not an internal sense of ‘__’, functions over the assertions of mental phenomena’s independent existence. (2)

The other late-nineteenth century route for pluralists under the flag of convenience of Concept Art could have been Husserlian phenomenology. The ur-clash between Frege and Husserl did happen: the latter required a faith in the continuous collective associations of and for mental acts and the former required, in brief, the following of functional rules. There is still much to be learnt from this divergence. (3) Neither route was without its agonies. There could be an assumption that the Fregean view that sentences contain varying degrees of complex propositional forms could provide a model of AL practice un terms of paradigmatic token and type; e.g., AL practice is a collection of sentences S which have functional arguments F(x) for a proposition p, etc. It could be assumed, and this would need a lot of methodological scaffolding to hold it up, that such a model would require a rule of truth conditions such that any p was true or was consistent within some truth condition rules. This problem was prior to the somewhat obvious condition that AL practice was dialogical. It will become apparent that semantic maps have formal problems nothwithstanding this dialogic aspect of AL practice, but there were theoretically intrinsic concerns about the properties of a model, such as truth conditions, satisfiability, what stood as the primitive items and so on for a model of (referential) meanings. It was assumed that a model or models would contain post-Fregean metaphilosophical conditions for thoughts and related speech acts/texts as having a propositional form of some sort. A large stock of formal connectives and other operators were given within the formal models. This was not a consequence of a fetish for a formal language system (of AL or for AL practice), nor did it represent a desire for AL practice to become a sort of purity-at-heart Principia schematics out of the confusions of a community of natural- language speakers. These formalities of the semantic problems did draw our attention, however, to some foundational problems of the How and the What of making sense and meaning (and it did link up with a rich source of other discourses). These observations on the relations and interests between formal and natural languages can be applied to complexities of the formal systems that were appropriated later and internalised within AL. Two cautionary qualifications are needed here. Firstly, the commitment to the Frege-style model of meaning was not instrumentalist. There was a prior commitment to a realism, to reflecting reality, etc. There was not just a pragmatic utility. And secondly, the model required the jury-rigging and the construction of variations of some distantly related contemporary theories. It was not an exercise of professional philosophical competences.

The referential failures of the propositional logic bemused some. Remember this is an Edwardian problem (Russell and so on) and alarming for an avant-garde gang to be involved in. Such classical examples as the proposition “The present King of France is bald” perhaps suggested that such failures would be tantamount to a prohibition on sentences/utterances having any referential sense, or that all that would remain would be a logical relativism. The formal structures as models and the interrelations between utterances and their speakers (as propositional) were assumed to capture the internal conditions of and the commitments to the representational realism of meaning for each and any of a community’s productive participants. Such modelling was not Scientism and not a Vienna Circle type of project to banish meaningless utterances (or expel the utterers); such ambitions would require instrumentalities fraught with a banal and violent absurdity. What was created was an embarrassment; i.e., that the semantic model proposed would hypothesise Kantian noumenal conditions: that the logic of thought qua meaning is the adopted model in-itself.

It can be assumed here that the sense of any utterance S within AL (as a practice of (a) Natural Language) would have reference, would satisfy truth and other formalisable conditions within a possible world within and of AL. This much was and is a matter of conventional semantic construction. (4) It also admitted the existence of coherence in the relations between utterances and also within a possible set of utterances (discourse/practice). There are two aspects of semantic modelling (after Tarski) that had some interest for us: (1) that sets of utterances (or sentences) may not have or share a consistency within a sets of sets that might be dubbed Art-Language Language and (2) that alethic modalities can be introduced and substituted by other modalities within the formal systems of propositional logics. The first aspect is related to paradox (sets of sets): the denumerability of sets and the conditions which are required for membership inclusion and consistency (i.e., the David Hilbert hat-trick of own goals scored against the formal systems). The second aspect was a given that modal operators in conventional formal systems such as possible and necessary can be replaced by permission and obligation (deontic logic), believe and know (epistemic) and tense logics. These two aspects of incompleteness and complexity combined in an interest on our part in the functional operations between utterances: the implicature. In the literature, the problem of referential success was that saying something about the world conforms to truth conditions (and their modal relatives) in a language L, and that the conditions are iterated. The distinction between the extensional and the intensional domains, therefore, disappeared. These domains were co-functional, and extensional commitments were paired with the intensional commitments and there was symmetry. (5) (The issue of ‘materialism’ in its varieties, historical and dialectical, as a uniform or monological ontology, was expropriated by the functional dynamics of ideology mapped indexically; the Marxian interests for AL were The Holy Family and The German Ideology as well as Capital.)

There was a realisation that the depth or the ‘compactness’ of meanings entailed a complex of modalities attached to sentences not envisaged by the semantic model makers, and that the belonging of sentences, or fragments thereof, to sets (of a language L[AL]) would remain formally incomplete or inconsistent, or both. Any aim to impose a propositional logic would collapse under indeterminancy. The propositional value of a sentence is therefore crudely expressed as the relation ‘any sentence at a specific location belongs to a specific language set of AL and this amounts to the conditions of its meaning’ (or Si Î (L) [AL] = p). This maintained an earlier commitment to the basic primitives of blurting (utterances/sentences) and implicature: meanings happened in the way sets of utterances were nested or embedded in the discourse/practice L [AL]. This also seemed intuitively attractive as a depiction of operators between utterances and speakers; we learn from utterances that belong historically to a set of meanings by putting such utterances into sets of meanings. (Learning is the function of changing semantic maps of sets.) This is not associativeness (i.e., Si ‘is like’ Sii therefore they both belong in the meaning set {p}) but rather a concern to display the reflexive rule-following. That’s just avoiding the problem of identity where a possible set {p} contains p & ~p. Propositional meaning was thought of as rule-following and the operators between utterances as transformational and generative. Hence the problem of the reflexivity included its resolution by building and changing the rules of following (and Going on).

As Charles Harrison has recorded, it was suggested within AL conversations that a ‘logical implosion’ occurred in AL practice during the Index projects (6). The term is somewhat vague and misleading, however. It suggests collapse or heat death when indexing was a way out of the paradoxes of incompleteness (of meaning), a search for the missing mass of practice rather than an absurd idealisation and an offensive closure requiring a brutish control over AL practice. The complexity of functional operators that you could push into the indexes was recognised early on: ‘Now systems of rules are of basic importance; the thing is that there are so many logics to consider’. (7)

Another view from the high-cultural demi-monde was that the Index projects and the related conversations were an impenetrable private language. AL’s view, most of the time, was that such a charge was a self-contradictory derogation uttered by users of a privileged ‘private language’. Opacity is, however, a substantive problem a priori, partly as a hermeneutic of intersubjectivity (I-Thou and Martin Buber and all that) and partly as a technically informed problem of quantification into intensional contexts. That is, language is both radically incomplete and translatable. Another AL view, some of the time, about the Private Language Problem was a scepticism deeper than the Cartesian cogito which had no relation to the cultured whining; that is, how do you know you are following the meaning in a (presumably well formed) language L, and which rules are you following to know you are following the meaning? Indexicality was an escape route from scepticism shut-down by non-phenomenological transcendence of absurdity; indexing actualises meaning. End of digression.

To return to complexity. The modal operators taken from the ‘so many logics’ available function in two important ways. They remove the first-order limits and also act as connectives between sentences. This was the case with the modalities of ‘propositional attitudes’ from the epistemic ones (x believes that p) to the dispositional ones (x loathes that p). Indeed, the propositional, in the sense of a sentence meaning, was the product of the complex relations of (a) modal operators, (b) a sentence or a fragment there of, (c) a set containing consistent membership rules and (d) a language set holding weaker or non-consistent rules of membership. This roughly outlines the semantic terrain, but two problems emerged: that the relation of (c) to (d) was regressive, and that functions (a) over (b) determining membership of (c) and (d) were iterative actual occurrences, in some intuitive sense, at time ti. If propositional meaning was considered desirable, and it was in the simple sense of reference to a world (or worlds) outside the utterance/sentence, then it could be captured as the membership and relationships to neighbouring members of a set of meanings. Again, iteration arose because (a) and (c) can occur in real time and (b) can be considered a textual primitive without historical conditionality. There is some irony in the complexity of the logical frameworks of operators and set membership and the simplicity of the primitives in the given relations of blurted concatenated utterances.

The meaning of a sentence (or fragment) was not an interpretative commentary of the AL practice, since it was a recursive function of AL practice. If indexicality was a pragmatic resolution of the problem of an absurdity or regress which possessed natural-language features intuitively, then we had been indexicalising all along. (Natural languages are indexical anyway.) Pragmatically, the natural language(s) of AL practice could contain the nomological features of meaning discussed above and maintain fuzzy boundaries between indexed sets of discourse within any (formalised) language L (AL), and so on. Indexicality entailed that a set of sentences would have meanings which generate and transform practical rules of a pragmatic logical form which change and continue within the discourse/practice. It was the case, therefore, that the quantification of a sentence (what meanings related to other meanings might exist within AL?) is highly sensitive, complex, specific and recursive vis-à-vis historical sets of meanings. Quantification into the intensional contexts without vectoral implications (i.e., what is projected through logical space and time) entailed conditions of meanings under two forms of reification: either a black hole of solipsism with its attendant privileges of singularity, or a ludicrously bureaucratic hegemony of protocol.

A non-cognitive problem was also intrinsic to the Index projects. Indexicality seemed to belong to a set of dispositional approaches which did not historically (or dispositionally!) belong to the traditions of formal systems (of logic and language), and yet they could be incorporated. Some of the modalities used to generate indices of meanings can be seen to belong to a range of other traditions: e.g., scorn, suicide/resignation, commitment and irony. (8) This was no accident. What was unusual was the realisation that the ‘Gödelian problem of not being able to count’ can be answered formally as existential commitments by historically specific meanings. Existential commitment – as quantification, or non-technical and traditional commitment – to meanings appeared as a natural type of Turing Machine which produced a pragmatic solution to incompleteness and non-closure. There were contingent conditions for the Indexes to operate the way they did which determined the character of the formal displays, but Turing Machines could take many forms. (9) The limits were temporal ones together with a given set of modal operators marking the indices with non-cognitive substitutes for (conventional) modalities. From the actions of indexing you could construe that a string of primitive syntactical items belonged to a set I of L {AL}, and the other sets indexed interrelated sets in specified ways. Remember, the modalities operated as the connectives and as the quantifiers of set membership. The non-cognitive aspect is important in that the given modality list (e.g., ‘Dead Parrot’, ‘Skeptical’, ‘Jobless and Gaga’, ‘“Dark Green”’ ‘Reinforce...anaphor’, ‘Resignation (sigh!)’, etc.) was a non-exhaustive list of AL attitudes.

The textual terrain offered was somewhat desolate: barely a texte at all. (10) The fragments of sentences given for indexical operations, which were contingent and not arbitrary, had only to be assumed to be actualisingly concatenatory to exert a semantic ‘pull’ towards each other. The minimum conditions of a formal language and/or Turing Machine were necessary. You moved one symbol/letter/word/box on and on again, depending on what the technological conditions (computer, pen and paper, etc.) would allow. What was strange was that Bx Cal – i.e., blurted iterative item x belonging to a set {al} – via routine indexing operations could create complexity, subtlety and depth. It was recursive and complete as a set-theoretical map of discourse during the indexing.  And this strangeness continued: the vectors of the indices of Bx Cal are the conditions of consistency and connectivity between Bx and By Cal, and there was symmetry in that Bx & By Cal described the logical space for Bxal.

Much has been left out in this canonical approach. For example, the aspect of community, not as a collection of career ambitions in search of a coherent justification, but in the sense of a commonality which possessed historical continuities and discontinuities as pre-conditions of the formal concerns outlined above. There is also the question of the translatability of the indexical machinery to iconic formats as syntactical items – a primitive item as ‘sign’ stands in a quantificatory indexical relation to a set of meanings (as propositional, referential) in contrast to an ideational semiotic... Or the problems surrounding the observation that a historically specific interpretation is a function of interests; cf. Gramsci and cultural hegemony... Or those problems connected with the thought that ideology, and not deterministic hegemony, saturates meaning at the micro-rules level of actual indexing Bx al By Cal, and that thus ideology is also interchangeable with the macro-rules which effect relationships between sets of indices or other normative propositions. The naturlogik of indexing was non-foundational; any fragment was substantively a strategic interpretation of the whole: like a shattered hologram.

We started by considering the dumb epiphenomena claims which were made by some conceptual artists about ‘ideas’, ‘concepts’ and ‘the mental’ to the point where the AL Index projects established a practice which had two unusual features as art (in a gallery): (1) nobody could exercise psychological privilege over the meanings made and displayed as these were explicitly incomplete – the Turing Machine was transfinite; (2) the Indexes satisfied the conditions of being a natural language L (AL), within the restrictions imposed by up-to-a-point coherent instrumentalities. Because indexing was an actualising process, the Indexes were not a metalanguage and not a metapractice. They were not an explanation or guide to what AL thought; they were AL thought. The realisation was – and there were no revelatory scenes of “Astounding, Holmes!” – that giving up on distinctions between the inside and the outside, cardinality and ordinality, blurred a lot of boundaries. The projects were anti-foundational, anti-reificatory, anti-hegemonic, anti-epiphenomenal, anti-psycho-phenomenalist, and they made the division between ‘art’ and its production absurd. That list is not an update of Ad Reinhardt’s via negativa, and the indexing was (almost) never a sort of technicians’ happening; the production conditions could be deep and powerful enough to contain indexed semantic heir-lines for their meaning (or their own set membership) far beyond the locale of their generation. Membership in both senses – as meaning production and as meaning producer – was co-extensive with what could be counted in the possibilities of the sets {L} of [AL]. Some commentators have mistakenly thought of the Indexes as the monolithic monsters of some project to manage art production (see Postscript below). As a transformational and transcending method, the Indexes made an absurdity of that mystificatory hermeneutics which serially bifurcates studio/gallery, (artist’s) mind/commentary, producer/observer, etc., and which is necessary for entrepreneurial success. The Indexing projects were shadowed by some dignifying adult ambitions to stabilise the historical meanings of AL, and far from coincidentally the meanings of a non-finite range of other neighbouring non-AL practices. The Indexes must appear to them like a descent to a virus.



Notes



(1) An adventitious, opportunistic and corporatist grouping of primitives which included ejaculated fragments of thought, fabricated propositions, just ‘had ideas’, a commitment to an aesthetic mental life, declarative utterances, self-referential sentences, exhortations, technical constructs and other epiphenomena. All were the products of non-circumspect and naive psychologistic realism, which is to say they were a form of anti-realism.

–––––

(2) This step function may partly explain the ambivalence of talking ‘in, about, as, by, with, for, and frequently against Conceptual Art.’ Art & Language, ‘Recollecting Conceptual Art’, ICA London, 1995, printed in this journal.

–––––

(3) For example, in art talk, the ‘Materialism and Aesthetics’ claims made in the 1980s by, amongst others, Sebastiano Timpanaro, Peter Fuller and Raymond Williams (!). This case, more sophisticated but as wrong-headed as the lumpen idea artists, proposed the neurophisical locations as the prime movers of meaning. (There was some other genetic-nurturing material added on.) But feelings, expressions and other mental events interpreted as electro-chemical occurrences cannot stand in for the meaning of (e.g.) “I feel depressed”, or any other propositions, even if you could locate the unique occurrence. It was a gain-saying, regressive mess. See (e.g.) Raymond Williams, ‘Problems of Materialism’, New Left Review, 109, pp.3-17.

–––––

(4) See A. Tarski, ‘The Concept of Truth in Formalised Languages’, in Logic, Semantics, Metamathematics, Oxford, 1956; and ‘The Semantic Conception of Truth and the Foundations of Semantics’, Philosophy, and Phenomenological Research 4, 1944 - e.g. ‘“Snow is white” is true if and only if “Snow is white is true” in Language L’ – but without Tarski’s views on semantic closure and the consistency of natural languages.

–––––

(5) ‘...to understand a sentence of first order logic is to know its interpretation in the actual world... [rather] to speak of different possibilities concerning our “actual” world than to speak of several possible worlds.’Jaakko Hintikka, ‘Semantics for Propositional Attitudes’, in Reference and Modality, ed. L. Linsky, Oxford, 1971.

–––––

(6) Charles Harrison, Essays on Art & Language, Blackwell, Oxford, 1991.

–––––

(7) T. Atkinson & M. Baldwin, ‘Documenta Memorandum (Indexing)’, in Documenta 5, Kassel, 1972, and in ‘The Index’, in The New Art, Hayward Gallery, London, 1972. Cf R. L Martin ‘but instead of the truth value of each sentence in L... each element of a set of points of reference (let I be the set of such points) is taken to have as much complexity as we require for an intended interpretation’. ‘Some Thoughts on the Formal Approach to the Philosophy of Natural Languages’, in Pragmatics of Natural Languages, ed. Y. Bar-Hillel, Dordrecht-Holland, 1971.

–––––

(8) David Wiggins considers attempts to escape from absurdity through non-cognitive commitment like ‘scorn or defiance (Camus), resignation or drift (certain orientally influenced positions), various kinds of commitment (R. M.Hare and J.-P. Sartre) and what may be the most recently enlisted member of this equipe, which is irony’. Truth, Invention and the Meaning of Life, British Academy, 1976.

–––––

(9) Beware of anachronisms here. A. M. Turing’s work on undecidability was known at the time, but his contributions to the first computers at Bletchley Park during the war and later at Manchester University were official (war) secrets in the 1970s. See ‘Computing Machinery and Intelligence’, A. M. Turing, Mind, 1950. Computers used for random generation of AL Index permutations in 1975 were industrial, the size of a house, and took three days to print off.

–––––

(10) Consider the case that the Index methodologically contains an anticipatory refutation of Deconstructionism, in that one texte indexically requires a quantificational commitment within the interpretation; the syntactical bleakness is below Derridean requirements and the semantic richness is greater because of the embedding of formally undecidable sets of meanings. Similarly, a problem of establishing an ordinality of textes seemed not to the point.