My aim in this series of blog posts is to consider the extent to which the knowledge-first epistemology can solve what I have been referring to as the unification challenge viz. how and why do the distinct priority theses endorsed by knowledge-first views come together to form a coherent and explanatorily unified framework for contemporary epistemology? In my second post, I argued that the unification challenge comes in two forms: one which concerns the connection between the metaphysical and conceptual aspects of the research programme (the inter-reading) and one which concerns the explanatory connections between priority theses within each family of claims (the intra-reading). There, I argued that despite first appearances, the latter is more problematic for the knowledge-first research programme: in order to show that the knowledge-first programme is genuinely innovative and distinctive, the knowledge-firster needs to provide an account of why the metaphysical priority claims are connected. [At the very least, I argued, examining the extent to which the knowledge-first programme can be explanatorily unified is a worthwhile endeavour].
The metaphysical ‘‘priority theses’’ which prompt the second unification challenge are as follows:
M-2 Knowledge is a relatively fundamental feature of reality
M-3 Knowledge does not have belief etc. as proper parts.
M-4 Knowledge is a mental state or phenomenon.
M-5 Knowledge plays important causal roles.
In this post, I’ll argue that unification challenge with respect to these metaphysical knowledge-first theses is best resolved by adopting a version of naturalistic epistemology - that is, the claim that knowledge is a natural kind. In my last post, I outlined the recent ‘pragmatic’ theory of natural kinds in philosophy of science due to P.D.Magnus (2012):
Pragmatic naturalism: A category k is a natural kind for domain d if the following conditions are met: (1)[The success clause]: k is a part of a taxonomy that allows scientific enquiry into d to achieve inductive and explanatory success and (2) [The restriction clause]: any taxonomy that excluded k would not do so.
As I outlined, this implies that natural kinds are always relative to a specific domain of enquiry. That is, on this view we should deny what is known as the ‘simpliciter assumption’ viz. the claim that there is a fact of the matter whether or not (for any natural kind candidate X) X is a natural kind (Goodman 2014, Bird and Tobin 2015,). Instead, the PM view is that natural kinds are always domain relative: natural kindness is not a one place predicate’ (‘electron is a natural kind’) but is rather a relational notion which relates natural kinds to specific domains of scientific enquiry (‘electron is a natural kind for the domain of particle physics’).
And what of the metaphysics of natural kinds? Pragmatic naturalism endorses a pluralistic claim about the the ontological constitution of natural kinds. For special science domains, Magnus argues that it reasonable to assume that the features of natural kinds (which explain and allow them to play this epistemic-explanatory role) is Boyd’s homeostatic property cluster view which I outlined in the previous post:
1. Property Cluster: natural kinds [in the special sciences] are connected to clusters of properties that regularly occur together. None of these are individually necessary for the kind to obtain and do not stand in a strict part-whole relation to the kind.
2. Homeostatic Mechanism: a causal mechanism underlies and explains why the clustered properties occur together.
Causal Import: natural kinds, [(1-2) explain] feature in important causal inductive generalisations.
In my view, this modified theory of natural kinds provides an attractive framework for contemporary epistemology. Crucially, I think it does so via explaining and unifying many of the metaphysical priority theses (M2-5) at the heart of the recent knowledge-first research programme in epistemology. That is, this modified naturalistic epistemology is a promising candidate for a framework which can provide the knowledge-first programme with explanatory unity. To see how this works, consider the following claims which are plausibly implied by the pragmatic naturalist thesis that knowledge is a natural kind:
Knowledge (and not mere true belief) plays an indispensable causal and explanatory role within various domains of inquiry.
Knowledge is a real, natural property with independent explanatory importance. It is, unlike the other epistemic categories, epistemically fundamental to certain domains of enquiry.
Knowledge will is a natural kind for some domains, but not all.
In those different domains, the properties associated with knowledge may be different. [ that is, this leaves open the possibility that in e.g. cognitive ethology ‘knowledge’ picks out RTB without claiming that knowledge is RTB simplicter].
These capture M2-M5 in the following way. The first should be straightforward: M5 is accounted for via endorsement of (1) [which is just a restatement of the same claim]. Claim 2 of the modified naturalistic epistemology gives the ‘relative fundamentality’ thesis a new, attractive meaning. Instead of appealing to metaphysically-loaded notions such as grounding, the knowledge-firster can claim that the sense of fundamentality is, at heart, epistemic indispensability: on this view, knowledge is fundamental in the same sense as the category ELECTRON, SPECIES and MEMORY are fundamental; they occupy a privileged place in our taxonomies of the world which secures warrant in their existence. In my next post, I suggest that this feature of naturalistic epistemology provides an answer to the value problem of knowledge.
What about the knowledge-first thesis that knowledge is a mental state which does not have belief as a proper part? (M3-M4). Here, the application of the pragmatic naturalistic epistemology becomes less straight forward. This results from the fact that for the pragmatic naturalist, the ontological constitution of the kind (in this case, Knowledge) is a secondary, a posteriori project: on this view, we must first look to find which categories are epistemically indispensable to a given domain, before subsequently examining which features of the world explain this role. This being said, I think there is reason to be optimistic that the naturalistic epistemology outlined here can capture the primary intuitions and motivations behind the knowledge-first endorsement of M3 and M4.
Let's start with M4. I assume that it is uncontroversial that many of the domains in which we would expect knowledge to be indispensable for will - like the example of cognitive ethology Kornblith discusses - be domains in the special and social sciences such as developmental and other forms of psychology, etc. That is, domains whose paradigmatic natural kinds are homeostatic clusters of higher level e.g. mental properties. If this is correct, it is likely that we can expect knowledge to have a HPC ontological structure. That is: it is reasonable to expect that in these domains, knowledge - if indispensable - is to be identified with various mental properties which in turn explain the kind's explanatory indispensability within that domain (M5). Similarly, this naturalistic view provides the knowledge-firster with a plausible explanation as to why the post-getter project in epistemology failed - the motivation behind the endorsement of M3. Here, the knowledge-firster (on this naturalistic construal) can cite a well known feature of HPC kinds viz. that the cluster properties which characterise the kind, while jointly sufficient for the kind, are not individually necessary. That is, on a HPC theory of higher level kinds, it is built into an account of the kind (Knowledge) that there will be various instances where the kind obtains, but not all of the typical properties which characterise the kind are not present (e.g. LIFE). The naturalistic epistemology which combines Magnus’ and Boyd’s theory can thus explain why knowledge is not in fact factorisable; that is,why producing an extensionally adequate, reductive account of knowledge, will fail.
As you no doubt will have noticed, the primary implication of this view - which sets it apart from Williamsonian views - is fundamentally a contextualist account of knowledge. On the view outlined in this post, Knowledge is not a kind simplicter but, a rather a natural kind relative to certain domains. Some may think that this is too far a departure from Knowledge-first epistemology to be worth considering. However, if the most plausible account of naturalism has contextualism built into to all scientific kinds, then I think, it is unreasonable to ask for more here. On this view, knowledge is as real, fundamental, and theoretically important as any scientific kind can be said to be. In my final post, I’ll argue that this view has further attraction viz. by offering us a plausible account of the value of knowledge which is lacking in current debate.