Goldman-on-the-Gettier-cases

Goldman, A. (1967). "A Causal Theory of Knowing." The Journal of Philosophy, 64(12), 357-366 and 369-372.

Goldman focuses on the second Gettier-cases, which involves the disjunctive claim either Jones owns a Ford, or Brown is in Barcelona (Goldman refers to this proposition as P)

“Notice that what makes P true is the fact that Brown is in Barcelona, but that fact has nothing to do with Smith’s believing P. That is, there is no causal connection between the fact that Brown is in Barcelona and Smith’s believing P… The requirement of such a causal connection is what I wish to add to the traditional analysis.” (Goldman)

Goldman agrees that the Gettier-cases undermine the traditional analysis of knowledge, because they satisfy all three conditions and are not genuine cases of knowledge. The modification to the traditional-analysis-of-knowledge that Goldman want to make is that there needs to be an appropriate causal connection between S’s belief that P and the circumstances that actually makes P true, in order for it to count as a case of knowledge.

Formally put, S knows that P if and only if P is causally connected in an appropriate way with S’s believing that P

This forms a causal analysis of knoweledge.

Causal connections to knowledge

Gettier outlines causal connections that are necessary in order for a belief to count as a case of knowledge:

case of perception

when the belief x is caused by the very fact that makes x true

Take the proposition “the cat is on the mat”. The fact being true causes certain sense-data that in turn causes you to form the belief

$$\text{fact}\to \text{perception}\to \text{belief}$$

case of memory

when the memory belief x is causally connected to the very fact that makes x true

suppose the cat is on the mat at time $t^1$ and at some later time $t^2$, you have the memory that the cat was on the mat at time $t^1$

Memory is understood as a genuine process of recall, so there is a causal chain between the memory belief that “the cat was on the mat at $t^1$” and the very fact that makes it true. $$ \textrm{fact} \rightarrow \textrm{perception} \rightarrow \textrm{belief (yesterday)} \rightarrow \textrm{memory belief (today)}

##### knowledge based on inference Goldman is uncommitted on this stance but posits that inference at least may constitute a causal connection. Goldman speculates that inference itself should count as a kind of causal process, by which new beliefs are produced from old one.

 \textrm{fact (?)} \rightarrow \textrm{perception} + \textrm{background belief } \rightarrow \textrm{inference}

Goldman gives the following example: imagine S perceives that there is a solidified lava in the countryside, and concludes P: "a nearby mountain must have erupted long ago". S's perception of the lava supports the belief in conjunction of S's background belief that lava us caused by volcanoes; S is making an inference P based on both the perception and his background belief. Suppose that S's belief that P is true. According to Goldman, to sat whether or not S knows that the mountain erupted depends on whether or not the eruption played a causal role in producing S's belief. So there are two possible ways: 1. *S does know*:  the lava that S now perceives was actually caused by an eruption. So, S’s belief, in term, is actually causally connected to the relevant fact. It’s this causal connection that ensures that S knows that the mountain erupted. 2. *S does not know*: though the mountain did erupt long ago, at some point afterward a man cleaned up the area by removing all the lava. Then, a century or so later, a different man wanted to make it look like a volcano had erupted, and he distributed solidified lava at various places in the area. Although S’s belief is actually true, the thought that makes it true has no causal connection to S’s belief, because whatever connection it might have had was undermined by that person that actually cleaned up the lava so long ago. ##### Knowledge of the future Consider the example: T tells S that he has the intention to go downtown tomorrow. So, S forms the belief P: T will go downtown tomorrow. The problem arises of how could an event or fact that has not occurred yet have any causal role in producing a belief. Goldman answers that to say that a fact is causally connected to a belief does not necessarily mean that the fact directly causes the belief. In the example, T's intention that caused S to form the belief also causes the fact of T actually going downtown tomorrow. In other words, T's intention is a common cause that establishes an indirect causal connection between the fact and the belief.

\textrm {T’s intention} \rightarrow \begin{cases} \textrm{belief (today)} \ \textrm{fact} \end{cases}

**Pattern 2 cases of knowledge** = these are cases in which the belief and the fact which makes the belief true are connected by a common cause. This can be applied to a variety of knowledge, like knowledge of the future, knowledge from testimony, etc. #### causal justification Goldman points out a significant difference between the causal analysis and versions of the [[Kim-on-justification-in-epistemology|traditional-analysis-of-knowledge]]. > “In at least one popular interpretation of the traditional analysis, a knower must be able to justify or give evidence for any proposition he knows. For S to know a proposition, P, at time T, S must be able at T to state his justification for believing P, or his grounds for P. My analysis makes no such requirement, and the absence of this requirement enables me to account for cases of knowledge that are wrongly excluded from the traditional analysis.” The difference between the two views is the definition of [[justification]] that traditional analysis of knowledge relies on an **internalist** theory of knowledge while Goldman's view relies on an **externalist** theory of knowledge. See *[[justification#internalist-and-externalist-theory-of-knowledge|Internalist and externalist theories of knowledge]]* Goldman posits that the justification condition is an inappropriate requirement that ends up excluding legitimate cases of knowledge, and his theory of knowledge is *externalist* in the sense that it doesn't require the knower to have cognitive access to the grounds of their knowledge. E.g. = knowing Abraham Lincoln's birth year might be knowledge one gains on the testimony found in a history book. Suppose that S retains the belief of Lincoln's birth year but forgets the basis for forming the belief. This is still a legitimate case of knowledge, even if one can't recall the justification for forming the belief. This would pose a problem for the traditional analysis, but not for Goldman's version, as the belief just has to have the right causal connection. Externalism liberates us from the requirement of being in a position to provide justifications that we have actual access to internally, and so liberates us from some of the problems of [[Skepticism]]. > “It should be noted that I have made no attempt to answer skeptical problems. My analysis gives no answer to the skeptic who asks that I start from the content of my own experience, then prove that I know that there is a material world, a past, etc. I do not take this as to be one of the jobs of giving truth conditions for S knows that P.” For an externalist, it does not matter whether the person can defend a belief against skeptical challenges, that is not necessary in order to have knowledge. Externalists emphasize that knowledge is about the objective situation of the belief relative to the reality. This view of justification can be understood in terms of reliability in the production of true knowledge and is called **Reliabilism** or **Reliabilist theory of justification**. This view is developed by Goldman in later writings. #### Comparing with Naturalism Goldman's causal analysis of knowledge is similar to [[Quine-on-naturalized-epistemology]] that it sidesteps skeptical arguments rather than refuting them. In both theories, we are free to study knowledge in causal terms without questioning our scientific and common sense beliefs about the world. Goldmans does not commit himself to the idea that causal processes involved in knowledge are based on input/output like [[Quine-on-naturalized-epistemology]] does. [[Kim-on-justification-in-epistemology#kims-critique-of-naturalized-epistemology|Kim's critique of naturalized epistemology]] is that it loses the normalized dimension. While Goldman's view also rejects the internalist idea of justification as a basis of normativity, the causal view still appeals to the notion of appropriate causal connections as a means to differentiate between beliefs that meet the standard for knowledge and those that don't.