Rejection of Individualism: Revealing the Social Foundation of Testimony

An argument given from the perspective of Frederick Schmitt.

How do we come to accept that today is our birthday? Or, that with the help of a couple thousand Athenians, three hundred Spartan warriors held hundreds of thousands of Persian’s for a few days at the Thermopylae? The answer is that someone tells us. But can we blindly accept anything that someone else tells us, or must we discover for ourselves if what was said is the truth? What lies at the heart of this matter is a question if conditions of knowledge require and are upheld by conditions that are social or individualistic. As we will see, by evaluating the role of testimony, justification at times requires social conditions.

To start we should clearly explain the difference between socialism and individualism. A social condition is one that requires more than one person. On the other hand, an individualistic condition is the negation of a social one.  An individualistic condition is one that requires only one member of a given society. The issue between individualism and socialism arises at the dependence of any knowledge condition on social conditions.

Let us now reflect on an argument for the best possible individualistic view of testimony’s basic dependence to social conditions. To do this, we must examine which strength of individualism has the best chance of being vindicated. Then, see how the argument would follow.

Strong individualism is the belief that nothing can be justified outside of the self. This is quite hard to follow because we cannot imagine a sudden realization of our name, date of birth, who our parents our, etc.  These must all be given to us by testimony. We can only move forward by accepting these as testimony and rejecting strong individualism.

Weak individualism is the view that we can listen to testimony for some justification, but that we ultimately must verify the validity of such justification ourselves. There are three forms of weak individualism that exist in an inductive, an a priori, and a coherentist version. The inductive version is the belief that we can accept justification from testimony as we hold testimony trustworthy, but that this trust comes from induction correlation caused by first hand observation of the testifier. The a priori version is mnemonically obvious. It holds that testimonies are held a priori. Finally, the coherentist view is that testimonial beliefs are justifiably caused by their coherence with non-testimonial beliefs. Of these, it appears as if the inductive claim prevails as the most robust, and has the best chance of defending itself against the Socialist view.

Historically, the idea that justification of testimony should be checked by perceptual first hand observation arrived from Hume. However, many things have changed since Hume’s time. For example we no longer have a reason to hold perception any more evidential than testimony (359). Just as we check our sources of information, we check the skill of our visual perception at an eye doctor.

Another reason to see why inductive weak individualism should be rejected is that we do not have enough perceptual experience to provide a foundation to support the reliability of testimony of which we can justify intuitively. There is no amount of experience that could enable me to reliably accept the testimony of another as justification. Furthermore, it is not natural to check the justification of testimony through perceptual experience. If it were so, humanity would not get very far into specialized topics like quantum physics. Much of our understanding and reliability biases come from testimony at some lower level. For example, in the perspective of the inductive weak individualist, I cannot understand what Jupiter is unless I have seen it. But for me to do this I will have to have a telescope. And when I go to use this telescope am I to trust that it works correctly? Such trust in a telescope seems to resemble the trust of testimony. So for the inductive weak individualist, they will have to learn about optics to discover it works just to see Jupiter. Such methodology does not appear to be what human beings, or any other creature, do.

In the case of children, they must accept that adults, caretakers, parents, etc. are reliable. If they do not, again, humanity would not get far. In the end children trust their parents and accept what they say as reliably justified to be true.  Only after children address people who they do not trust do they skeptically check the validity and reliability of others testimony. And it is most likely the case that this is done by asking those that we do trust. In this process, the reliability of testimony increases until an idea of credibility is understood. In these processes there is little first-hand observational checking of validity that takes place.

The inductive weak individualist could then turn to argue that if the justification process of testimonial reliability requires only a few good instances, then it is okay to believe we have enough experiences to accept the reliability of testimony. However, this argument is twisted and is does not liberate the inductive weak individualist’ position.

In certain situations it is okay to take one instance a justification for a belief. But in these instances underlying beliefs hold a required foundation that allows us to be justified. For example, our belief that there exist general properties among items of the same type allows us to apply a given property to all such items of that type.

Hilary Kornblith, however, takes a different direction. Kornblith holds that small sample inductions can be acquired reliably through inferences, so long as the property, which is being attributed to, is of a natural kind. Natural properties work because they seem to have properties that correlate well with one another. Because this correlation exists, one can reliably infer that natural properties exist outside of this small sample. Furthermore, one does not have to believe that a natural kind exists. All that is necessary is that one does exist.

If Kornblith is correct, does this apply to the testimony of beliefs? If so, the inductive weak individualist could use this as a defense. For this to happen, they would need to treat the desired property in a small sample induction to the reliability of testimony as the natural property of the small sample induction, so that a small sample induction to reliability can be justifying. The reason why this induction from a small sample is reliable is because of its being well correlated. Therefore, in this defense, it is necessary that the said desired property be well correlated.

An obvious rebuttal to this kind of justification can be seen in accidental situations. Therefore, something else is needed, like a “non-accidental” feature, where the process we use cannot be invalidated by accidents.

When we take this two-tiered argument it looks similar to that of Goldman. Goldman distinguishes between these reliability processes. He believes there are natural processes and acquired processes. The natural are processes which we do not acquire though learning; hence a process that we naturally have. So how then do we obtain reliable acquired processes? Goldman suggests that there is a metareliable proper method-selecting process that does this for us.

From here we can derail the inductive weak individualist view.

It looks like a metareliable process is required for acquired processes and some natural processes. Furthermore, a natural process may not be reliable just because it is justified. 

As we will see, the need for a metareliable process can be won over by a simple reliablism, as long as the selection process were correctly combined with the selection methods and processes so that a single process be reliable just in case the selection process is metareliable and the selected method or process is reliable.

The requirement for a metareliable process is fulfilled for a small sample induction because we do inductions on natural kinds. Children do in fact rely on or choose natural kinds when doing a small sample induction. These children, who categorize objects in natural kinds, use a method called focused sampling, where they examine objects for properties and if they find that these properties are independent, yet share a strong relationship with the object, other natural kinds are more likely to be found. Because of this evidence, we may believe using this two tired method, tat selections made by a selection process is evidence for a metareliable process.

We must then be required to use the same process for small sample induction of natural kinds to the small sample induction of testimonial beliefs. First, The induction must be reliable. Second, the testimonial property must correlate well with the testimony’s reliability. And third, the induction must be metareliably selected. We will see how the inductive weak individualist will fail this third requirement.

If the way in which the inductive weak individualist argues is the best way for their belief to defend itself against socialism, we will see how the induction of testimony is not metareliable. The analogy between the induction of natural kinds and that of testimonial belief is not applicable. To see this, we can treat the testimony as having natural kinds. The property of testimony and its being well correlated to the reliability of the property does not allow for focused sampling to work, because there are only two properties for the sample. Furthermore, these two sampled properties are not independent of each other. If I wish to check with the reliability of testimony from a testifier, I must check with the testimony. This dependence reveals that the case for metareliable processes on natural kind inductions does not work with that of testimony. And therefore, that testimony is not a natural kind.

It is the case that some metareliable processes work on small sample inductions of testimony. However, this metareliable process is social, not individual. We can select testimony by a metareliable process, however, this requires prior beliefs that undoubtedly come from someone else (this makes it social). This must be the case, because I cannot test the testimony of my mother explaining that (as Roderick Chisholm puts it) ‘what appears to me bluely’ is in fact blue. Such a test would be impossible unless I accepted social inductions.  But this would defeat the individualist’s purpose.

The question is then what social process allows for the development of selection processes? Naturally, it is trust. Children trust their caretakers. From these caretakers/parents they develop and learn what to believe. The processes of selecting testimony that he caretaker finds adequate for the child is then what the child learns. Surely there appears to be an evolutional understanding as to why this makes sense. The child trusts the person who keeps them alive because it must be good, for they are still alive. This is a social in the fact that it is a social situation that the guardian of the child cares for the child, and that this is a social relationship. If this were not the case, our world would not resemble what it is today. And it is metareliable because the process is outside of the child’s control and it is reliable.

One could challenge this by saying that this is not a metareliable process but just gullibility. They may also claim that gullibility is not a social condition. This would, however, be an incorrect assessment. If it were gullibility and not trust, what is to change? It could be that trust is itself the same as gullibility. Then gullibility would just replace trust in the theory, where it stands as metareliable. Furthermore, it is social, because the child participates in this arrangement where they believe another. Moreover, we must use social conditions to discover gullibility is the process that a child relies on.

As I have revealed, by refuting the best possible individualist defense of testimonial belief, we see that testimony is a social condition which knowledge requires. Because we do in fact use testimony for our knowledge at some base level, knowledge is social.

Now the question arrises, how can we improve knowledge by looking at social aspects of life?