All Embracing But Underwhelming…

Goldman, Alvin, ‘Knowledge in a Social World,’ OUP, 1999

p.5 - Book is an essay in social veritistic epistemology – an essay in the social practices that have a favourable impact on knowledge (as opposed to its contraries of error (false belief) and ignorance (lack of belief)).

p. 6 – is a normative and not merely descriptive account.

p. 10 – six standard ‘post-modernist’ objections to truth explicated.

p.21 - ‘In a significant sense, then, only the world confers truth and falsity. There is, nonetheless, something correct in constructivist claims. The thought contents we consider and accept are rarely “given” to us by the (nonhuman) world. They result from our own biological resources and linguistic activities; in that sense, they are human constructs or products. Since knowledge involves belief, and belief is in contents that are so constructed, there is merit to the claim that knowledge is (partly) a social construct. But since knowledge is true belief, knowledge also involves truth; and what is true, as we have seen, is not a human construct as opposed to being of the world. Hence, it is wrong to say that knowledge is merely or entirely a human construct. Similar points pertain to postmodern claims that knowledge is always “local,” “contextual,” or “situated” rather than timeless or universal. Knowledge partly consists of belief, and belief is always local or situated because it is always the belief of a particular knower or group of knowers who live at particular points in time. But knowledge also partly consists of truth, and when a fully determinate proposition is true, it is true for all time, not just at particular times or places. The proposition that there is a cup on the kitchen table at such-and-such an address at noon, Greenwich Mean Time, October 18, 1997, is either timelessly true or timelessly false. Similarly, the proposition that African slaves were brought to the Americas between the sixteenth and nineteenth centuries is true for all time. Its truth value does not change as a function of the locale of a given believer. In this sense, truth is not something local or situated or socially constructed.’

p. 23-4 – Distinction between SuperStrong and Strong Knowledge. Strong Knowledge is knowledge under some kind of forcing condition where we do not have to be able to exclude all possible situations (evil demons et al); only relevant objections to A knowing P need to be considered.

p. 24 – Book is not on Strong Knowledge (some kind of JTB concept) but on Weak Knowledge (true belief). Goldman claims this is what people want.

p. 24-5 – Puts forward his notion that Weak Knowledge (W-Knowledge) is all we need to talk about by claiming that we use know in this weak sense all the time. But it seems that W-Knowledge, only being true belief, presents problems. One example Goldman uses is Seth and Beth. Seth believes there is extraterrestrial life and Beth does not. It seems Goldman is implying that the fact of the matter (that there are, or are not, ETs) means that one of them has a true belief and thus knows. This doesn’t seem like the term we would use here. Goldman is willing, however, to admit that W-Knowledge might be a term of art/technical term.

p. 42 – Correspondence Theory of Truth – a proposition is true if it corresponds to the facts.

p. 52 – Deflationary Theory of Truth – there is no difference in content between “It is true that P” and (the assertion of) “P.”

p 59 - modifies the Correspondence Theory of Truth to add in that the relation must purport to describe reality.

p. 60 – takes ‘true’ to be faithful; a fidelity notion.

p. 62 – “As long as anything makes a proposition true is part of reality–construed as broadly as possible—this fits the correspondence theory as forumlated by (DS) [descriptive-success].”

p. 69 – ‘People have interests, both intrinsic and extrinsic, in acquiring knowledge (true belief) and avoiding error. It therefore makes sense to have a discipline that evaluates intellectual practices by their causal contributions to knowledge or error. This is how I conceive of epistemology: as a discipline that evaluates practices along truth-linked (veritistic) dimensions. Social epistemology evaluates specifically social practices along these dimensions.’

p. 69-71 - Arguments against Consensualist Epistemology (that all we look for in a theory of knowledge is generating agreement).

p. 72-5 - Arguments against Pragmatism (the notion that a theory of knowledge should produce utility).

p. 75-9 - Arguments against Proceduralism (rationality criterions et al).

p. 79-82 – In arguing for veritism Goldman posits that some might argue that the selection of the best processes to acquire knowledge could be subject to the objection that there is no ideal way to select such processes. Goldman argues that this will be true of other varieties of social epistemology such as pragmatism et al and that if we talk about target practices (the practices we want to choose from as the reliable indicators of truth) and selection processes (the practices we use to choose target processes) then we can develop stories as to how we might converge on the best target practices (he uses the distinction between majority voting by experts, consensus views of experts and weighted voting by experts to show that, over time, we might converge on weighted voting as our preferred practice).

p. 84 – notion of ‘superpractices;’ vision is a practice and sense perception (as a collection of practices) is a superpractice. Often discussions of circularity will centre on practices vs. superpractices; sense perception could be viciously circular but visual sense data might not (because it could be backed up by touch, et al).

p. 85 – Some of the concern for circularity comes from the notion that a practice can be used to backup the same practice; I see a tree and I check that by looking again at the object. This isn’t necessarily bad, though, as I might not get the same sensation; every instance of the practice can be treated as an independent test.

p. 83-7 – Arguments against circularity in re veritism.

p. 87-94 – In developing the veritisic conception Goldman further distinguishes between fundamental veritistic value (states such as knowledge, ignorance, error et al) and instrumental veritistic value (practices that promote or impede the acquisition of fundamental veritistic value).
Goldman also uses a degree of belief schema and a trichotomous schema for fundamental veristic value (believe, reject or withhold judgement).
In adopting the veristic model of valuable states Goldman uses a question-answering model; states of belief have value when they are responses to a question that interests the agent. Thus the range of beliefs that the agent doesn’t hold that are not of interest to her have no negative impact on the veristic value of her belief states. The veristic value of an agent should always be assessed in relation to what interests her.
Goldman’s model focuses on changes over time; a practice that increases fundamental veristic value over time is value more so than one that affects no change and is better than a practice that reduces veristic value.
The focus on practices not only applies to current practice but also to future practices; we want to be able to assess whether a new or future practice will be better or worse than a current one. Thus we talk not of frequencies but rather of propensities.
We can also talk about absolute versus comparative evaluation; absolute looks at the improvement a practice makes all by itself whilst comparative looks at the improvement a practice makes in relation to a rival.

p. 94 – ‘Interest’ can be defined in re individuals and in re institutions.

p. 95-6 – The notion of interest is moderate (rather than pervasive) in Goldman’s schema; because an agent might be motivated by non-epistemic factors for a particular question we should not place too much weight on interest, although practices that do no nothing to improve an agent’s core interests we should also be suspect of.

Chapter 4 – Testimony

p. 105 – Assuming that the speaker is truthful and the audience credulous then reports that will have the greatest change in veristic value are ones that interest the audience, the thing being reported is not (usually) predictable and the report must not be stale (i.e. these three conditions seem to satisfy the definition of a newsworthy event).
Goldman is interested here in the great change in veristic value, so the most effective change in veristic value is when a low degree of belief changes to a (relatively) high degree.

p. 110 – Goldman asks what acceptance (of testimony) practice we should seek in his social epistemology. He goes for a satisficing practice, one that might not be the best but will show improvement in veristic value. This is an absolute rather than comparative mode; we look for the practice that, on average, gives us an increase in veristic value rather than the practice, comparative to others, that will give us the best improvement in veristic value given a particular context (this answer will be Bayesian).

Chapter 5 – Argumentation (but Goldman treats this as an extension of 4; some testimony is more than factual reporting (it can be argumentative))

p. 132 – Arguments, to Goldman, are more than simply two sides stating their reasons for a belief; the speakers must engage (endorse or criticise) with each other.

p. 136 – the role of conversation as being informative.

p. 153-4 – argues that, in re Critical Thinking, sometimes we should take into account the arguer as well as their argument when deciding what to believe…

p. 185-6 - Links to Chomsky.

Section 8.4 - Rebutting the Theory-ladeness of Observation

Section 8.5 - Rebutting the Underdetermination Thesis

Section 8.12 - Recognising Authority

–

Goldman, ch. 4

p. 103-5 - Four stages of testimony-related activity in re social epistemology: Discovery, production and transmission of messages, message reception, and message acceptance.

p. 105 - Our best acceptance practices are (likely?) to be those with the great change in degree of belief.

‘Under what conditions can prior DBs be expected to be low? When the following three conditions are satisfied: (1) the occurrence of the reported event or fact is not predictable from background information, (2) audience members do not themselves observe the reported event or fact, and (3) audience members have not already received reports of the event from other communicators.’

p. 121 - ‘Theorem (4.1): Suppose an agent observes the value of a certain variable (for example, the testimony of a witness), which depends probabilistically on the true answer to the question Q(X/-X). Suppose the agent uses Bayesian conditionalization to update her DB vis-à-vis the question Q(X/-X). Finally, suppose that her subjective likelihoods match the objective likelihoods; the likelihood ration ≠ 1.0; and her prior DB ≠ 1.0 or 0. Then the objectively expected change in truth possession (V-value) vis-à-vis Q(X/-X) is positive.’

p. 123 - Proposes three kinds of subjective evidence for udging likelihoods: reporter’s competence, reporter’s opportunity and the reporter’s sincerity/honesty.

p. 127 - ‘Notice that Coady’s challenge is a challenge about epistemic justification. It raises the question of how people can justifiably accept the testimony of speakers, not the question of how often they get truth from trusting testimony.’

p. 129 - ‘In its simplest form, justificational reliabilism says that a belief is justified if and only if it is produced (and/or sustained) by a reliable belief-forming process or sequence of processes. For a testimonial belief to be justified it suffices that the general process of accepting the reports of others mostly yields truths. Accepting the reports of others is really a specific pattern of inference, where inference is construed as a process that takes some beliefs as inputs and generates new beliefs as outputs. In the testimonial case, the inputs include beliefs of the form “Person R reports X” and the outputs are belkiefs of the form “X.” What is required for this process to be reliable? Assuming that hearers accurately represent speakers’ reports, what is further required is that those reports be generally true. So credulity acheives reliability if and only if it is exercised in an environment in which speakers’ reports are generally true. If this condition is satisfied, then the (simple) reliabilist theory of justification assigns the status of “justified” to testimonial beliefs, whether or not believers have the inductive basis for regarding testifiers’ reports are reliable. This is similar to the other theories I have been expounding, which also allow a testimony-based belief to be justified even if the hearer doe snot have enough testimony-free evidence for the general reliability of testimonmy. Of course, a sophisticate form of reliabilism would also accomodate “defeating” evidence, so that if the hearer has evidence against a testifier’s credibility, she is not justified in believing the testifier’s report.’

Ch. 5

p. 133 - ‘First, there is a link between justification and truth: if a premise belief is justified, it is likely to be true. The link between justification and truth follows from most theories of justification. It is straightforward under reliabilism, because a reliably produced belief is produced by a process that usually outputs truth.’

(i.e. faculties ‘designed’ to be verististic.)

p. 145 - ‘(TEP [Truth-in-evidence Principle, a form of the total evidence principle]) A larger body of evidence is generally a better indicator of the truth-value of a hypothesis than a smaller, contained body of evidence, as long as all the evidence propositions are true and what they indicate is correctly interpreted.’

p. 145-6 - TEP asserts (Goldman assumes) that conforming cases are more common than nonconforming cases.

–

Goldman’s model of an argument.

p. 134 (Monological)

(1) the speaker believes the asserted conclusion;
(2) the speaker believes each of the asserted premises;
(3) the speaker is justified in believing each of the asserted premises;
(4) the asserted premises jointly provide strong support for the conclusion;

p. 136

(5) At least some members of the audience to which the argumentation is addressed do not already believe the asserted conclusion (the fewer such members believe it, the better);

p. 137

(6) All the premises presented in the argument are credible to at least some members of the intended audience (the more such members, the better);

p. 138

(7) The premises-conclusion relationship is displayed or explained in a fashion that promotes its comprehension by the audience;

p. 139

(8) The audience has no defeater for the argument that the speaker endorses;

p. 140 (Dialogical)

(9) Critical argumentation may challenge or rebut a speaker’s argument by either (A) presenting a defeater, (B)) denying the truth of some premises, or (C) denying the strength of the premises-conclusion relationship. Such criticism or rebuttal is generally veritistically beneficial, so long as it conforms with the preceding conditions, especially (1)-(4);

p. 141

(10) Rebuttals of a speaker’s argument should be accurate.

(11) More effective rebuttals of a speaker’s argument are to be chosen over less effective rebuttals;

p. 143

(12) When a proponent’s credal or epistemic situation changes as a result of criticisms of his argument, he ought to offer a public retraction of his earlier claims (especially when there is a listening audience);

(13) In selecting responses to critics, a justifiably undissuaded proponent should weigh both the perceived seriousness of the criticisms by the audience and the audience’s comparitive receptivity to the potential responses available to him;

p. 144

(14) When there are existing or foreseeable criticisms of one’s main argument, a speaker should embed that argument in an extended argumentative discourse that contains replies to as many of these (important) criticisms as is feasible.