Here are two observations to prepare the way for the answer to this question that I want to propose. Firstly, formation of beliefs by uptake from perception and testimony are among the most basic transactions in our cognitive lives. There is absolutely no reason to think in advance that accounts of how the resulting beliefs are justified will be very like the accounts that might apply to more sophisticated undertakings – development of scientific theories; arrival at conclusions by abductive or inductive or deductive inference. Secondly, after several thousand years of trying to develop accounts of justification in basic cases that do echo the accounts that seem right for more reflective belief formation, we have still not managed to explain how uptake from perception and uptake from testimony confer justification on the resulting beliefs.
I suggest that under these circumstances the appropriate methodology, when asking how justification by uptake from perception and testimony work, is not to treat these cases along lines already developed for more reflective belief formation. Rather, we should try something else: look at how formation of beliefs by uptake from perception or testimony actually works; look for a kind of positive normative status that the resulting beliefs might be claimed to have; then ask whether this kind of positive normative status might count as theoretical justification (justification for belief).
This is the methodology that lies behind the upgrade from practical to theoretical justification in the account of perceptual demonstratives from last time. Our ‘that’ beliefs have a kind of positive normative status: they are formed by a route that is guided by the mind’s need for cognitive focus. Can we claim that these beliefs are (therefore) theoretically justified? Well, if we do grant this claim, the 1-5 line of thought from last time goes through. And if this line of thought goes through, we have an account of justification for perceptual demonstrative beliefs which meets the most fundamental (and perhaps the only) test for theoretical justification: theoretical justification is truth conducive; if a kind of positive normative status for beliefs is a kind of theoretical justification, beliefs having this status must be reliably true.
And now we can also address the circularity worry which has been rattling around for the last few days. Fixing Reference is committed to the following three principles:
Principle connecting aboutness and truth – An <a is F> belief is true iff the object it is about is F. (If my belief that Jack has fleas is about my dog, it is true iff he has fleas.)
Principle connecting truth and justification – Justification is truth-conducive: if your belief is justified, you will be unlucky if it is not true.
Principle connecting aboutness and justification – A body of beliefs is about object o iff o is the unique object whose properties the subject will be unlucky to get wrong in justifying the beliefs the way she does.
I take the first two principles as given, and use them to argue for the third. The circularity worry arises if we read the principles as explaining truth in terms of aboutness; aboutness in terms of justification; and justification in terms of truth. But that is not how the Fixing Reference proposal goes. I am treating each of the three conditionals as having explanatory significance. But I am not treating any of them as explaining one term in terms of the other. In particular, I am not treating the truth and justification principle as explaining justification in terms of truth. Rather, I am treating it as stating a condition that a kind of positive normative status must meet if it is to count as theoretical justification: the upgrade from a kind of positive normative status to theoretical justification is available only for varieties of positive normative status that are conducive to truth.