Date: Sun, 22 Mar 1998 16:37:32 -0800
Subject: HAB: Re: Differentiations of understanding: beyond logical atomism




Joe,

Your recent remarks make much clearer to me why someone might believe
that bootstrapping (or presuppositional “performative contradiction”) is
going on in Habermas’ position. But I don’t think there’s anything
“murky” in Habermas’ thought. Your own view of things (apart from
Habermas) is very interesting, but “compositionality” is an approach to
linguistic cognition that has met with much dispute from “modularist”
and “connectionist” researchers in cognitive science. I contend that
Habermas’ work has more affinity with cognitive science than linguistic
relativism (despite common readings of Habermas as a linguistic
relativist).

Relative to the one statement from Habermas that you quote, “without
getting into details,” you write that Habermas “describes his aim as one
explaining how ‘validity claims would have emerged from the integration
of the narrower, paleosymbolically anchored concept of moral authority
with the other components of speech acts’ ([TCA-2:]72).” But this is not
Habermas’ general aim, in part of his critique of Durkheim from which
you're quoting. There, Habermas is making the point that illocutionary
force cannot be derived strictly from presumptions of normative
validity.

However, I agree with your assertion--which the above-quoted passage
from you is taken to bolster--that “[t]he emergence of propositionally
differentiated speech occurs in social evolution with the integration of
the three prelinguistic roots of communicative action....” But, if you
believe that these roots of communicative action *are* prelinguistic (as
Habermas does), then your later comments on compositionality loose
relevance, as does, by inference, your inferentialist reading of things.

To Habermas’ mind, the prelinguistic roots of communicative action are
world-relations. Within this increasingly differentiated simultaneity of
world relations (first as self vs. world, then self vs. other, self vs.
object, and other vs. object), ontogenic “simultaneity” of prelinguistic
and linguistic relations undoubtedly *pertains* to the development of
consciousness, along with essentially nonlinguistic features of
intelligence which co-constitute modalities of mind that are relevant to
communicative representation. In all this, though, Habermas never claims
to *reduce* the differentiation of world relations to components of
speech acts. Rather a *homology* of ontogenic differentiations (world
relations and components of speech) are implied by our *analytically*
given model of speech. This model articulates an isomorphism between
language and world which arises ontogenetically. Habermas nowhere claims
that linguistic development creates world differentiations; rather world
differentiations and propositionally-differentiated speech arise
together.

>From a compositional point of view (which seems to me to derive from
logical atomism), it would be plausible to believe that “the integration
of these three [prelinguistic] components is what gives speech acts
their illocutionary-mode/propositional content structure, *and* what
transfers the understanding of speech acts onto the basis of
criticizable validity claims.” But this is not Habermas’ view (and a
compositional view is not largely supported by researchers in cognitive
science, especially inasmuch as computational neuroscience tends toward
connectionism, rather than inferentialism). Nowhere does Habermas give
paleosymbolic life a “componential” character. Cognitive development
grants differentiations, so to speak, which
propositionally-differentiated speech implies. It is differentiated
world relations that are primarily implied, and criticizable validity
claims are a largely post-conventional development of non-criticizable
world relations, by grace of our innate disposition to discover ever
more sophisticated differentiations (in levels as well as of kinds).

Accordingly, it is not primarily the case that “understanding validity
claims involves knowing how they can be discursively redeemed,” except
to post-conventional understanding. Primarily, understanding involves
knowing what one means, which implies why you believe or say something
(a child eventually comes to realize), which becomes differentiated into
that which we distinguish as warrant and backing. But persons can
provide warrant long before they can provide backing, and even a skilled
provision of backing can fall within a conventional approach to
rationality that does not yet demand or offer *argumentation* for one’s
sense of relation between warrant and backing (like young teenagers),
let alone *discursive* (hypothetical-testing) argumentation (like young
adults), which may be variably competent.

It is not the case that “Habermas considers mastery of at least the
rudimentary components of argumentation to be [a] necessary step in the
development of linguistic competence.” One can believe, as I do (and
claim Habermas would agree) that, contrary to your point, one *does*
“learn a language first, then learn how to use it to argue with people”
and still believe, as you do (and I agree) that “children learn
inferential relations early.” But the issue is (1) what kind of
inferential relations are available early on and (2) do acts of
cognitive *representation* (emergent prototyping) ontogenically precede
acts of inference for any given level of inference? (1) Post-Piagetian
research has not undone the basic Piagetian discovery that sensori-motor
intelligence is the condition for the possibility of
concrete-operational intelligence, which precedes formal-operational
intelligence. Rather, controversy involves how early a differentiation
of levels emerges and how modalized such emergence may be. (2) The basic
developmental dynamic of decentration, which is basic to Piagetian
research and is pervasively accepted by Habermas, is not reducible to
relations of propositional differentiation. Quite the converse:
Propositional differentiation is an emergent property of cognitive
decentration. As ontogeny becomes increasingly linguistified,
differentiation is increasingly linguistic--often even linguistically
dependent, in given environments of communicative control. But a
reduction in principle of decentration to linguistification is
untenable, given the modularity of intelligence. Habermas’ focus on
linguistification, for a theory of communicative action, is not
incommensurable with a broader cognitive view which he has adapted to
his project. Ultimately, Habermas is not a linguistic relativist. This
can be shown in his formulations, prior to TCA, of a formal pragmatics,
in _Comm. & the Evol. of Soc._.

I agree with you that “Habermas's intuition is usually formulated in
something more like the following way: In order to be using
propositionally differentiated speech (as opposed to mere signalling),
you have to be able to distinguish between the different ways in which
your speech act can fail, e.g. whether the hearer rejects it because
what you said was false, or because it was inappropriate for you to say
it at a party, or whatever.” But I disagree with your following claim
that “Habermas claims that in order to get this, you have to be able to
put yourself in the position of an *arguing* opponent....” Basically--as
Habermas shows in his work on moral-cognitive development--one has to
put oneself in the other’s *perspective*. If the other has an *argued*
disagreement, then you can’t understand their contested validity claim
unless you understand the argumentative stance the other takes. But the
capacity for role reversal developmentally precedes the
formal-operational capacity to take up argumentation, as is evident in a
child’s learning to see things from their playmate’s point of view, and
simply *thereby* obtain a sense of perspective, independently of
warrant, or backing, let alone argumentation. From an adult perspective,
we would give preference to a warrant that has good backing over a
warrant without any pretext of backing; and we would give preference to
the backing that can be argued well over the backing that cannot be
defended cogently (as is the case with dogmatism).

Nevertheless, I agree with you that Habermas believes that “the
different components of the speech act are distinguished by the
different kinds of validity claims they raise, and understanding the
validity claims means understanding the reasons that could be give for
or against them.” But this is irrelevant to the issue of how basic to
cognition are adult inferential relations. An inferentialist reading of
early development does not accord with Habermas’s sense of cognitive
decentration. I should add that Piaget was more inferentialist than he
should have been, as indicated by post-Piagetian research (see
Sternberg, cited in my previous posting to you). But this inferentialism
is not what Habermas wants from Piaget; rather the cognitive dynamic of
decentration is the point. And understanding of meaning
post-conventionally cannot be validly applied to a child’s point of
view.

You write that “Habermas clearly *came up* with the position that
understanding speech involves knowing how its associated validity claims
could be discursively redeemed before he read Dummett, but he didn't
actually have an *argument* for this position before that time.” This is
false. His basic position was evident in the mid-1970s, even before the
essay on “universal pragmatics” (CES), which is evident in _Legitimation
Crisis_ (1973), as well as his 1970 essay “Toward A Theory of
Communicative Competence” (in _Recent Sociology 2_, Hans-Peter Drietzel,
ed., New York: Anchor, 1971).

Though it may seem to be the case that “the compositionality
requirement, which is what motivates Dummett's position, is the only
thing that provides any reason to believe that knowledge of meaning
*consists of* knowledge of justification-conditions,” this is irrelevant
to the developmental issue that I was counterposing to your association
of Habermas and Dummett in my previous posting (as well as presently).
*Adult* knowledge of meaning as justification-conditions must be the
basis of discursive claims about the intimacy of meaning and truth, in
reconstructive discourse. But the content of that
reconstruction--genealogy of the understanding of meaning--must be
differentiated from the discourse of reconstruction itself. Children
have a sense of warrant in their earliest sense of meaning, but this is
very embodied and familial, not justificatory at all. The
presuppositions of argumentation in communicative action is a
presupposition of world relations and lifeworld embodiment of emergent
differentiations between meaning and warrant, meaning and  defensible
warrant, meaning and arguable backing, etc., *within* the preceding
emergence of differentiations between self and world, etc., whose
cognition is increasingly linguistic, but not primarily so.

Whatsmore, I believe that it is *not* the case that the compositionality
requirement is all that provides reason to believe that knowledge of
meaning consists of knowledge of justification conditions. In any case,
it is not the case that meaning consists of knowledge of justification
conditions. Rather meaning necessarily *implies* knowing what truth
conditions count; but meaning altogether involves relations of self and
world--subjective, intersubjective, and objective world relations--which
are not simply defensive.

Intentionality is not reducible to representability. Habermas doesn’t
reduce understanding to the redeemability of validity claims. Rather, he
poses the *rationality* of understanding in terms of the redeemability
of validity claims. But even for rationality, it can be argued (as some
in cognitive science do) that the compositionality requirement is not
all that provides grounds for truth. Cognitive perspectives which I
cited in my previous posting in this thread imply a “connectionist” and
embodied sense of the *emergence* of inferential performance from
“natural kinds” of mental categories (not necessarily mentalese. A
philosophical perspective here that I find very attractive is the short
book by Hilary Kornblith, _Inductive inference and its natural ground:
an essay in naturalistic epistemology, MIT, 1993).

It’s not developmentally true, outside of claims about *adult*
competences, that “the rules of inferences (understood in the Toulmin
sense, i.e. pragmatically) are what give you the substitution relations
that determine the contribution that words make to the meaning of
sentences in which they occur.” What “determines” meaning
reconstructively--Toulmin’s context--is a different *kind* of matter
from what determines meaning developmentally. Projecting adult
perspectives onto developmental beginnings does indeed foster the
confusions that led to Ken’s dilemmas.

“People have all kinds of fancy theories,” you note, “but the basic
objection to perceptual relations [as the basic connection between
utterances] has been that they don't give you compositionality.
Inferential relations, on the other hand, do.” But this kind of point is
only relevant synchronically, so to speak, at a developed level of
reflection. It’s irrelevant to a claim about the basic connection to
world relations which Habermas claims utterances have. The precedent
development of differentiations which speech embodies emerges from
cognitive accomplishments that pertain to the entirety of intelligence,
and the notion of perception can encompass the holism of embodied
representation better than a notion of speech, as Merleau-Ponty,
cognitive science (in the emergent, connectionist vein), and holistic
developmentalists (such as Howard Gardner, who is greatly influential in
American educational thought) have argued, as argued in publications I
cited last week.

Cognitive psychology is, I would argue, far beyond the
representationalist paradigm that Wittgenstein abandoned. I agree with
you that “[t]he way the pieces of a sentence fit together is just not
the same as the way perceptual objects are configured.” But this is
irrelevant to the issue of how sentences may come to have pieces in the
first place. A compositionality argument presumes the cognitive
capacities that grant the elements of holistic experience AND THE
CONSTITUTIVITIES which are allegedly decisive for composition. The
compositional argument does not support semantic holism, developmentally
conceived.

Consequently, the dilemma of performative contradiction--which is
fostered by your view--belongs not to Habermas’ work, but to a tradition
of logical atomism that isn’t relevant to Habermas’ project.

Best regards,

Gary



     --- from list habermas-AT-lists.village.virginia.edu ---