Charles Taylor, Philosophical Arguments (1995)
Questions & Answers #5: "The Importance of Herder"

Augustine or Adam's (in Genesis) view: "Words get their meaning from being used to designate objects," 80. This requires a correspondence model of epistemology. Condillac accepts this model to argue for a notion of how language originated.

1. On what basis does Herder reject Condillac's causal explanation for language?

"As an account of origins, it presupposes just what we want explained. It takes the relation of signifying for granted," 80.

Notice the AI (artificial intelligence with computers) parallel for this point: "The problem is that Condillac endows his children [cf. computers] from the beginning with the capacity to understand what it means for word to stand for something," 81. The background (where relationships lie) is unrecognized. "The possibility that some sophisticated robot . . . which matched us in the correlations of utterance to world [i.e., could pass the Turing Test] & yet might not understand anything, makes no sense . . . ," 82.

2. "To understand Herder's objection to Condillac, we have to take the inner standpoint, that of the agent," 82. Reductionism (from the paradigm model of Descartes mathematical physics) assumes that an objective set of descriptions can replace a subjective experience to better explain an event. What does Herder (T. & Pragmatism) say about this? Do they accept it?

They reject it: "we can't accept an account of how a creature possesses [uses] language [or any other human event] exclusively in terms of the correlations an observer might identify between its utterances, behavior, & surroundings," 82. If you want to know the meaning of the event, you must ask the person - thus entering into the context/culture.

3. At the top of p. 83 T. explains the "problem of induction," without identifying it as such. Can you identify this? & paraphrase it?

Paraphrase: selecting an item/atom from a background (Dasein), makes it foreground (a focal point). This attributes a name to the item/event. However, "acquiring this kind of understanding [induction] is precisely the step from not having to having language [syntax, relationships]. So it is just this step that a theory of origins would have to explain" & doesn't because it overlooks the act of induction/judgment, 83.

4. What would T. & Herder say to B.F. Skinner's behaviorialist, stimulus-response (Pavlov) models of human thinking? "What Condillac's children have to grasp in order to learn a new world is ____________ from what animals grasp when they learn to respond to signals.

"different," 83; the analogy is false. Human beings have a tacit knowledge of a background which unifies their experience into a meaningful, purposeful activity. When asked, they can usually induce the purpose & state it together with the context within which it is meaningful. Thus items/elements within the context are never atomic signals; they are concepts or instruments requiring a context/gestalt & goal/purpose.

5. What is the difference between a signal & a word (84)?

"To learn to use the signal is to learn to apply it appropriately in the furtherance of some non-linguistically defined purpose or task," 84. The point is that the context conferring meaning on the action is externally defined. The trained animal responds to a stimuli to obtain an associated reward. It cannot understand the meaning/significance of the signal as a symbol grounded in a context or outlook/theory. "It is a question of subjective understanding [meaningfulness], of what rightness [conformity with the paradigm/axiom logic that governs the system] consists in for it, qua what word is right," 84-5. This must be an insight, a realization, a judgment; not a behavioral response which is judged or evaluated by an observer - who must do the same thing, once removed.

6. T. writes, "But what is missing in the French thinker is any sense that the link between sign & object might be fundamentally different when one crosses the divide," 88. What divide does he have in mind?

". . . Humans can control the flow of their own imagination, whereas animals passively follow the connections triggered off in them by the chain of events," 88. For animals background must be co-extensive with perception. In contrast, humans can tacitly select elements from perception & memory to construct gestalts/backgrounds. This is very close to defining discourse communities.

7. What is the "fault [or error] in any designative theory of meaning," Adam designating names for entities, 89.

"what the background provides is treated as though it were built into each particular sign [atom], as though we could start right off coining our first word & have this understanding of linguistic rightness [contextual appropriateness; conformity to system logic] already incorporated in it," 89.

8. T. paraphrases Kant's claim that discourse (consciousness) is unified (i.e., a fusion between focal points/atoms & background/context) on the top of p. 91. "Coming through Hume, it [designative theory] held that the original knowledge of reality came in particulates, individual impressions. At a later stage the bits were connected together" in innumerable ways/theories, 90. Kant (followed by Pragmatism) believes this model is flawed. How so?

"Kant . . . argue(s) that this relation between knowledge [the atomic bits] & object [what it is about; how it is meaningful] would be impossible if we really were to take the impression [the input of atomic bits] as utterly isolated, with no link to others [no pattern supplied by Dasein, by human contexts in which we have experience]. To see it as about something is to place it somewhere, to give it a location in a world that has to be familiar [meaningful] to us in some respects," 91.

9. Explain how language is not additive (designative) but transformative. Consider how "retrieving the background and situating our thinking" work, 92.

Hume was wrong; words are not photocopies of perception. Words transform the status of perceptions by placing them in a context within which they are meaningful. Perceptions may be linked/associated in various ways. As Darwin told us, these links are not inherently meaningful; they are temporally proximate. Whereas "To possess [or know] a word of human language is to have some sense that it's the right word," i.e., to know the context within which it is meaningful, 93.

"Language is not something that can be built up one word at a time," 94.

10. What is the difference between feeling an emotion and expressing that feeling, 97?

This comes close to Freudianism: "feeling an emotion" implies that one does not recognize the context within which it occurs or has meaning. Because the background/context is operative (unconsciously) when it is not (consciously) recognized, feelings seem overwhelmingly powerful; they seem to entirely possess us or have us in their grip. By "expressing our feelings we can come to have transformed feelings," 97. By articulating feelings, we experientially place them in a context; they become parts of a whole, even when we cannot entirely (objectively) identify the whole. They become episodes or events in my life, further contextualized by recognizing the discourse community vocabulary that is most appropriate: e.g., elated on a date; despondent when I got my exam score; etc.

More quotes from "The Importance of Herder"
On To #6: "Heidegger, Language, Ecology"
Oct. 96