Pattern and Process

1. The Issue at Stake

Linguists, whether of the formalist or functionalist persuasion, are in the habit of speaking of grammar in quasi-processual terms, as if the patterns they describe and attempt to explain were themselves dynamic entities, whereas what they often are doing is in fact describing static relationships between the layers or modules of their theoretical frameworks which can be “projected” the one onto the other for various explanatory purposes. This has all too often resulted in a thorough ambivalence (if not downright confusion) as to what is best described as enduring linguistic pattern and what is best described as linguistic process. Much of the confusion is due to the conveniences of descriptive practice and the persisting influence of the computational model of thought, and can be avoided if one treats language patterns (individual grammars) in the manner of Whitehead as “eternal objects” that define the grammatical norms adopted by a community and treats linguistic process—the creative application of those norms to the communicative needs of individuals—as “prehensions.” This perspective is hardly revolutionary in itself, indeed it is quite reminiscent of the Humboldtian distinction between language as energeia—creative human activity—and language as ergon—culturally maintained product (Humboldt 1988), but its particular Whiteheadian formulation can help sharpen our understanding of the murky ontological basis of the Saussureian dichotomy between langue and parole (or, in its more syntactically defined Chomskyan guise, between competence and performance).[1]

What the Whiteheadian perspective more specifically offers modern linguistics is a handle on the psychologically real processes lying behind the deployment of language pattern in thought and communication (two sides of the same coin for Whitehead). By “psychologically real” I do not refer to the level of neurons (which has excited cognitively minded linguists in recent years, perhaps prematurely), but to that of phenomenologically meaningful experience, which must of course in turn be relatable—if only indirectly—to a more inclusive neural “nexus.” Many linguists otherwise interested in these matters still tacitly accept the digital computer analogy of the mind/brain, which precludes real-life meaning and intentionality. Even in more recent “neural net” and “connectionist” approaches to language modelling the mind/brain is simply a black box, whose inscrutable “hidden layers” are in themselves quite meaningless. The Whiteheadian perspective allows a degree of penetration into the workings of the black box in terms of his general theory of the conscrescence, in so far as this can be applied to neural networks.[2] It provides a smooth transition between low-level, analogue connectionism and higher level digitalized/symbolic processes and reflects the emergence of complex rational behaviour from simple cellular activity obeying basic physical laws. The successive steps of the Whiteheadian concrescence (where a simple conformal stage precedes more complex “aesthetic” and “intellectual” ones) suggest how distributed neural networks might function within thinking, communicating beings, with numerous processes occurring “at once” in real time although logically ordered and under mutual constraints. Moreover, Whitehead’s subtle theory of prehensions allows for the differentiation of what is often assumed by network theorists to be undifferentiated neural activity into a battery of specific prehension types that operate at hierarchically accruing levels of complexity. In this way, simple physical and conceptual prehensions (such as those involved in the recognition of patterns by specific areas of sensory cortex) are woven into propositional prehensions and, beyond this, into the most complex kinds of intellectual and judgemental prehensions. Much of this is relatable to what is in fact known about the workings of the hierarchically organized human cortex and the processes surging through it—including those of sub-cortical provenance.

Prehensions (that is, positive ones) are feelings, and feeling in its broadest sense lies at the heart of the Whiteheadian framework. Applied to linguistic memory—both semantic and episodic—it offers an alternative way to conceive of the storage and access of meaning (and its components) to the computer-based one involving labelled nodes or “addresses.” Furthermore, Whitehead’s theory of “subjective aim” deals with the dimension of internally-generated intentionality (and therefore of pragmatics) that is lacking from all existing computer simulations of linguistic behaviour. It allows us to see the gradual working out of the complex and largely indeterminate intentional dynamic that lies behind the production of individual utterances. This involves the gradual adjustment of a subjective aim towards its determinate “satisfaction”—its resolution (re)attaining maximally intense homeostasis—via specifically activated prehensions, e.g. of relevant parts of the “eternal object” of the grammar of English. The distinction between direct and indirect speech acts seems over-simplified by comparison, though Whitehead had little to say himself on specific grammatical ways of encoding complex linguistic intentions. I have attempted to fill out some of the missing detail in my own work.

2. An Interpretation of Whitehead’s Philosophy of Organism as Applied to Linguistics

In a monograph on what I have called the Pattern and Process (or Whiteheadian) perspective on linguistics I attempted to illustrate how some of the central ideas of Whitehead’s philosophy of organism can be applied with benefit to that broad and fragmented discipline (Fortescue 2001). The central point that I argued there was the importance of upholding two complementary perspectives, that of language as Pattern (in terms of conceptual “repeatables” that include the norms and socially sanctioned rules of grammar) and that of language as Process (in terms of individual occasions of use, and the prehensions and subjective aims of which they consist). In Whitehead’s universe, Process is complemented by Pattern, which is pure potentiality (although itself dynamic). Pattern only exists in so far as it is borne by actual occasions engaged in processes of concrescence. In any self-organizing system, one might say, Process is both the source and the result of Pattern, and this applies equally to language. Process is in general the passing on of feeling into expression, and feeling (or prehension) is the reception of expression (MT 23). What is in focus in on-line language use—as it is in perception—is dynamic, transitive action and what remains in the background is the more static context, largely presupposed and outside of conscious awareness. But without that static background—including, crucially, the constraints on the form of what can be expressed if one wishes to be understood—there can be no structured communication at all.

I presented in my monograph a cognitive interpretation of the Whiteheadian framework in relation to such areas as speech act theory and pragmatics, syntactic patterning, propositional semantics, language processing, information structure in written texts, language acquisition and historical change. An appendix contains a sketch of the logically successive stages in the concrescence of a single output utterance in English, indicating both the eternal object types and the prehension types relevant at each stage (Fortescue 2001, 251-53). In a number of articles since then I have attempted to apply this approach to somewhat more detailed linguistic matters of speech production, ranging from the simple phonetic level to the most complex intentional phases (see especially Fortescue 2007a, 2006 and 2007b).[3] Here follows a preliminary list over the types of “eternal object” relevant to language Pattern, as proposed in these works (taken from Fortescue 2007c). Whitehead himself, note, was no taxonomist (of eternal objects or otherwise), and was less than specific as regards the intended range or variety of these “forms of definiteness.”

A. Concepts
i. predicates (of activity, state, quality, etc.)
ii. logical subject types

B. Relational/structural patterns
i. morphosyntactic
ii. phonological
iii. (co)referential
iv. logical-relational
v. semantic

C. Types of subjective form
i. discourse modulation functions
ii. epistemic stance
iii. validatory attitudes

D. Intentional types
i. discourse acts
ii. strategies (and patterns of inference)
iii. behavioural norms
iv. background communicative principles

E. Nexus types
i. contexts of situation
ii. social/historical and textual embedding
iii. sign (e.g. word) types

F. Propositional types
i. simple
ii. complex

Compare the following list of the main Whiteheadian prehension types that I have used in my treatment of linguistic Process (here taken from Fortescue 2007a). About these Whitehead was quite specific, but the application to language is my own.

a) Physical. From the afferent (comprehension) perspective, this simple type of prehension—typical of the early “conformal” stage of a concrescence—is involved in the analysis of the raw speech “data” (phonological, prosodic, gestural accompaniments, etc.) and of relevant perceptual aspects of the communicative setting. Such conformal prehensions occur automatically at the neural level of the speech organs, not at the level of the conscious “personally organized nexus.” Physical prehension of the speech signal in the mode of “causal efficacy” leads via abstraction from the input data (by negative prehensions that isolate the signal from “noise” and allow for higher-level contextual variation) to the recognition of sequences of individual phonemes in the more fine-grained mode of “presentational immediacy.” This constitutes a bridge to the following, “supplementary” phase of the concrescence.

b) Conceptual. This kind of prehension is involved in ensuing phases of the concrescence where specific concepts are recognized as being associated with the speech input chain (via phonological form and symbolic reference links). From the efferent point of view of the speaker, corresponding eternal objects (“concepts” in the widest sense of recurring patterns) are prehended as potentially contributing to the satisfaction of the initial aim (the overall discourse goal). The lexical items evoked (along with their respective argument frames, etc.) are then integrated according to mutual compatibility with the general morphosyntactic patterns associated with the complex discourse act aimed at. Negative prehensions are needed in order to select for expression only the most suitable items thus prehended. In this way the subjective aim of the concrescence (inherited from the speaker’s immediate past) becomes more and more determinate and specific through successive conceptual and propositional prehensions.

c) Hybrid. Hybrid prehensions are the type involved in recognizing the intention behind another person’s utterances (by empathy and abduction) or one’s own previous intentions. They are also involved in accessing relevant presuppositions and making necessary “bridging inferences” in order to fill in inconsistencies in accruing “mental models” (the enduring objects of memory).

d) Indicative. This is the kind of physical prehension involved in acts of reference (to a logical subject—including those “labeled” by proper names), or of establishing a discourse topic (if there is a choice among multiple logical subjects). Such a prehension will highlight certain regions of the relevant nexus (or contextual network of relations) for potential predicative comment.

e) Propositional. This kind of prehension combines types (b) and (d) above, in the simplest case assigning a predicate to a singular logical subject. The “imaginative” propositional sub-type integrates the prehension of a predicative pattern with the “indicative” prehension of a logical subject that does not actually display the pattern—it assigns, one might say, a predicate to a logical subject in a relationship that is “imagined” rather than perceived in the immediate physical setting. A proposition as such, note, is neither an “eternal object” (a repeatable pattern) nor a nexus, but a hybrid relationship between logical subjects prehended within a nexus (as under d) and a predicate (an eternal object) of a kind that the human mind/brain is especially attuned to prehend. It is not necessarily expressed or articulated verbally. What is prehended (as with all higher level prehensions) is a species of contrast (between the nexus and the eternal object).

f) Intuitive judgments. This kind of complex prehension is a combination of (d) and (e) at a still higher level of complexity within the supplementary phase of a concrescence, whereby a whole proposition is grasped within its overall nexus (thus allowing the assignment—or suspension—of a truth value). Differences of “subjective form” here may reflect varying epistemic stances on the part of the speaker. The intuitive judgment forms the basis of processes of inference and abduction of various kinds.

g) Anticipatory. These “feed-forward” prehensions monitor the progress of the concrescence towards the intended satisfaction of its initial subjective aim, in particular as regards the effect on the hearer that the final utterance can be expected to have (this may involve general principles of communication or specific knowledge concerning the recipient and the given situation of context). The later stages of the specification of the “subjective aim” involve anticipatory prehensions.

3. Whitehead’s Impact on Contemporary Linguistics

In assessing Whitehead’s influence on linguistics one must bear in mind that Whitehead the holistic thinker was for a long time totally eclipsed in these quarters by his atomistic erstwhile colleague Bertrand Russell, who—via formal semantics in particular—was far more in tune with contemporary linguistic concerns. Whitehead’s larger scheme of things does not obviously connect either with formal (as opposed to “natural”) logic or with Wittgenstein and Ordinary Language philosophy (which arose by reaction to Russell and the semantic formalists). These were the principal areas of interface and communication between philosophy and linguistics in the years immediately following the war (cf. however Butt, this volume, on Whitehead’s influence on Firth). This lack of recognition is perhaps not surprising, given that Whitehead’s writings on language—apart from broad generalizations—is limited to his slim monograph on Symbolism (S) plus the corresponding section of Process and Reality, a work that makes such daunting terminological demands on non-specialists and is so permeated with metaphysical speculation that most hard-nosed linguists have shied away from approaching it directly. With the florescence in recent years of Cognitive Science (in its broadest, interdisciplinary sense), there are signs of a rekindling of interest in Whitehead on the part of some linguists, however, this time not in the formal symbolism of the Principia, but in the perspective on the ontological underpinnings of language production and comprehension that his full-blown philosophical system provides. It remains to be seen how far this will lead. In general, psychologists have been more open to Whitehead’s ideas than linguists have.[4]

Noam Chomsky, the most influential linguist today, is not unaware of Whitehead’s work, but he appears not to have appreciated his full potential significance for the field. In fact he began the book that introduced the notion of “Cartesian Linguistics” to the world (Chomsky 1966) with a citation from Whitehead, though this was only in connection with Whitehead’s approving attitude towards the “genius” of the sixteenth century and did not refer to his own philosophy. Now Chomsky has always claimed that his generative approach to linguistics is part of the “cognitive revolution” of the post-war years, but by this he is referring primarily to the “new” computer analogy of thought. It has become more and more difficult to see how the various avatars of his original generative theory have anything much to do with psychological reality. Only his reiterated—and untestable—belief in the Cartesian “innateness” of the language (read “syntactic”) facility remains constant. The principal difference between the perspective of realist Whiteheadian and that of arch-rationalist Chomsky lies precisely in the question of the ultimate anchoring of the language facility, both in individual cognition and in human evolution. For Whitehead it clearly has its roots in pre-linguistic cognition, in particular in perception, and is not an autonomous “device” that appeared suddenly out of the blue by some freak mutation.

Language for Whitehead is both symbolic and propositional in essence, but his term “symbolic reference” covers both perception and language. The proposition structures not only language but also perception (in the mode of “immediate presentation”)—the correspondence between “proposition” in this broad sense and linguistic “sentence” is far from direct. Language “elicits propositional prehensions,” i.e. invites the hearer or reader to adopt the “subjective form” of the proposition referred to, the whole intentional background of the utterance or sentence concerned. Symbolic reference, by coordinating perception in the mode of presentational immediacy (of sensory symbols) with that of causal efficacy (the holistically “felt” meaning or intention behind the act), lies close to the theory of meaning as “affordance” now widespread in psychology (cf. MacWhinney 1999, 216f.), but unlike the latter—basically a matter of distributed sensory associations—it is bidirectional (the intended meaning may also evoke the symbol).[5] What language introduces beyond the propositional prehensions of perception (and which raises it above the simple exchange of social signals) is a vastly increased clarity and intensity of experience, and this it does by liberating the speaker/hearer from mere sensory fact. It opens new vistas of abstraction and creative imagination. The primary function of language is to share our feelings about the world, not to communicate facts about the world (compare Franklin1990, 244). This is a definite advance on the Saussure-inspired “conduit” metaphor of linguistic communication, and represents a side of Whitehead that must have appealed to Chomsky and his Humboldtian insistance on the creativity of language. What the Whiteheadian approach promises, then, is an account of the “emergence” of language from pre-linguistic reality, one that does not rely on any unknown—and probably unknowable—”language acquisition device” in the manner of Chomsky himself. In this, Whitehead actually lies closer to the functionalists of today than to the formalists (despite Chomsky’s proprietary claim on the term “generative”). For Whitehead, the source of grammar is to be found in the subjective form of propositional prehensions which act as “lures for feeling,” and this is close to the heart of speech act theory. The most typical grammatical marking of subjective form is in the choice of mood, which reflects the speech act type, although it enters into many other aspects of intentionality and function that are reflected in grammar (including, for instance, adverbial modification).

This said, it must be admitted that there are nevertheless drawbacks to generalizing Whitehead’s philosophical system directly to speech production and comprehension, drawbacks mainly of a terminological nature. Nowhere is this clearer than in connection with his key term “eternal object.” Whitehead’s system needs interpretation—if not translation—to become accessible to mainstream linguists. This term, which is innocent in itself, since all that is “eternal” about such “forms of definiteness” is that the patterns are repeatable and therefore communicable. I have elsewhere suggested equating these with “concepts” in the broadest sense (including percept and intention types), though this may mean wrenching the purely cognitive aspects of Whitehead’s system out of its broader metaphysical context. I for one happen to believe that this is perfectly justifiable and results in an internally consistent perspective on cognition in general and on language in particular.[6] This may still not be enough to satisfy empirically-minded sceptics, however well the system seems to cohere as a purely psychological framework. Many would see Whitehead’s system as all too vague and general, even if “right” as regards the larger picture. It is certainly true that there are aspects of linguistics where a Whiteheadian perspective has little to contribute (one thinks especially of purely inductive linguistic typology). However, there is no reason that predictions deriving from the Whiteheadian “model” should not be tested empirically—and this is indeed a direction I am currently working on.

Another objection that may be levelled at Whitehead by linguists is his simplistic or at least “old-fashioned” understanding of the proposition as consisting simply of logical subject(s) and predicate. Information structure theory—not to mention various schools of syntax—have advanced far beyond that, it might be claimed, for instance as regards the distinction between “topic” and “subject” and the varying valency frames (both semantic and syntactic) of individual predicates that may require anything from zero to three arguments. In fact the broader contextual setting of a predication (which determines such matters of linguistic expression as topic assignment) is easily accommodated within the framework of Whitehead’s higher prehension types, and, as is well known, Whitehead made a point of promoting “polyadic” relations beyond the traditional binary structure of the proposition. However, the objection misses a more important point, since sentence type and “underlying” proposition are for Whitehead only very indirectly linked (here he sides with Chomsky—and with Plato). The Whiteheadian perspective supports the modern version of linguistic relativity whereby the sentence types provided by the language we speak may influence (but not determine) the kind of propositions we entertain. What is quite up-to-date about Whitehead’s understanding of the “proposition” is precisely that he treats this hoary abstraction in processual terms, namely as the combining of an indicative act (prehending the logical subject(s) of the proposition) and a predicative act (prehending an eternal object or concept, the affordances that constitute the contents of the predicative expression).[7] The propositional prehension organizes these into a higher order unity corresponding to a state of affairs, which may in turn be elaborated by a further prehension of the logical-subject-in-context that allows the assignment of a truth value (or other attidudinal indicator of subjective form). This is reminiscent of the way in which complex predications are built up in layered varieties of functional grammar. Far from being a hindrance, the centrality of the “proposition” in Whiteheadian thought may afford new vistas on the nature of linguistic processes.

Notes

[1] One of Chomsky’s principal calls to arms in heralding the transformational/generative “paradigm shift” of the late 1950s was an appeal to Humboldt’s emphasis on the dynamic, creative aspects of language a century earlier. This was part of his attack on the static, structuralist kind of linguistics that had prevailed within mainstream American linguistics since Saussure’s equally “revolutionary” replacement in the early twentieth century of the diachronic, historical approach to language by a synchronic one that focused on grammatical systems as such. It should be borne in mind, however, that Chomsky’s “competence” (like Saussure’s “langue”) corresponds to Whiteheadian Pattern, not Process. Dynamism for Chomsky is a matter of the mathematical generation (not the “on-line” production) of sentences.

[2] Which is precisely what I have argued for in Fortescue (2001) and elsewhere. It should be pointed out that recent “hybrid” versions of neural network models (“constraint satisfaction networks”) are now exploring the possibility of combining a “hard” symbol-processing capability with a “soft,” holistic connectionist architecture—see Pinker (2005, 13) for a brief discussion. This opens the way in theory for the modelling of higher level, proposition-based Whiteheadian prehensions within a neural “substrate” nexus.

[3] One specific area where an extension of the basic Whiteheadian perspective into areas of speech production has proved conceptually fruitful is that of the involvement of the “hybrid” type of prehension in the inheritance from actual occasion to actual occasion of “sub-routines” whose satisfaction has to be completed before a superordinate aim can itself be satisfied (Fortescue 2003). Another is the incremental building up of mental models from linguistic input (Fortescue, 2007b).

[4] See for example Damasio (2000, 226 & 287) for the relevance of Whitehead’s notions of “causal efficacy” to background feeling and of the non-continuous, “quantized” self for consciousness (the latter notion may in fact have been inspired by William James, who greatly influenced Whitehead).

[5] Thus one may say that the left temporal lobe connectivity around Wernicke’s classical language centre corresponds in Whiteheadian terms to the “faint” nexus pertaining between certain fine-grained auditory percept types in the mode of presentational immediacy (“symbols”) and widely distributed “affordances” felt in the mode of causal efficacy (their “meaning”). The nexus is “faint” in the sense that the relationship between the expression and the content of symbols is arbitrary. It is nevertheless deeply entrenched within the individual language user’s mind/brain. The participation of the frontal lobes is needed for this static nexus to enter into linguistic processes of any kind (interpretative as well as productive). The persisting internal dependencies between the twin poles of the symbolic relationship are bilateral (mutually determining), whereas linguistic processes as such (initiated by the left frontal lobe) are essentially in a relationship of unilateral dependence vis-à-vis the patterns they draw upon—they utilize whatever conventional patterns are available for moulding an intention into communicable verbal form.

[6] Note that Whitehead himself allowed for this interpretation by introducing his “category of reversion,” which is unnecessary if one prefers (as he himself did) to explain subjective aim in terms of hybrid prehensions of God (PR 250). Franklin, one of the few besides myself who has expanded on Whitehead’s theories concerning language, follows him in this (Franklin 1990, 39).

[7] This is the approach now taken in recent versions of Functional Grammar (cf. Mackenzie and Gómez-González 2004, for instance). It can be traced back at least to Searle (1969).

Works Cited and Further Readings

Chomsky, Noam.1966. Cartesian Linguistics (New York, Harper & Row).

Damasio, Antonio. 2000. The Feeling of what Happens (London, Vintage).

Fortescue, Michael. 2001. Pattern and Process. A Whiteheadian Perspective on Linguistics (Amsterdam and Philadelphia, John Benjamins Publishing).

Fortescue, Michael. 2006. “The non-linear nature of linguistic change,” in Competing models of linguistic change, edited by Ole Nedegaard Thomsen (Amsterdam, John Benjamins), 17-31.

Fortescue, Michael. 2007a. “The non-linearity of speech production,” in Structural-functional Studies in English Grammar, edited by Mike Hannay & Gerard J. Steen (Amsterdam/Philadelphia, John Benjamins), 337-351.

Fortescue, Michael. 2007b. “How to catch a mental model by the tale,” Acta Linguistics Hafniensia 39, 125-152.

Fortescue, Michael. 2007c (forthcoming). “Eternal objects, figurae and memory,” in Memory and Language, edited by Cornelia Zelinsky-Wibbelt (Frankfurt, Peter Lang, SABEST).

Franklin, Stephen T. 1991. Speaking from the Depths: Alfred North Whitehead’s Hermeneutical Metaphysics of Propositions, Symbolism, Language, and Religion (Grand Rapids, MI, Eerdmans).

Humboldt, Wilhelm von. 1988. The diversity of human language structure and its influence on the mental development of mankind (Cambridge, Cambridge University Press). [Original German version 1876-1880].

Mackenzie, J. Lachlan., and María de los Ángeles Gómez-González (eds.) 2004. A New Architecture for Functional Grammar (Berlin & New York, Mouton de Gruyter).

MacWhinney, B. 1999. The Emergence of Language (New Jersey, Lawrence Erlbaum).

Pinker, Steven. 2005. “So how does the mind work?” Mind and Language, 20, 1, 1-24.

Searle, John. 1969. Speech Acts: An essay in the philosophy of language (Cambridge, Cambridge University Press).


Author Information

Michael Fortescue
Department of Scandinavian Studies and Linguistics
University of Copenhagen, 2300 Copenhagen S
fortesq@hum.ku.dk

How to Cite this Article

Fortescue, Michael, “Pattern and Process”, last modified 2008, The Whitehead Encyclopedia, Brian G. Henning and Joseph Petek (eds.), originally edited by Michel Weber and Will Desmond, URL = <http://encyclopedia.whiteheadresearch.org/entries/thematic/language/pattern-and-process/>.