An Introduction to Phenomenal Conservatism

Phenomenal Conservatism (PC) is a foundationalist theory of justification that can be applied to perception as well as the a priori. Michael Huemer formulates PC like this:

PC: If it seems to S that p, then, in the absence of defeaters, S thereby has at least some degree of justification for believing that p (Huemer 2007).

PC takes seemings to be the epistemically relevant mental states. Seemings are appearances that something is the case, such as the appearance of a desk in front of me. It seems to me that there is a desk in front of me. Seemings are propositional attitudes: it seems to S that P is the case. For it to seem to S that P, the proposition <P> must be the content of S’s seeming. But just because seemings are propositional attitudes, it doesn’t follow that they lack a phenomenology. Seemings have a feel of veridicality; they present their contents as if they were the case. In other words, seemings have assertive content. Contents that are presented to the subject assertively have a phenomenology of, for lack of a better descriptive term, truthiness.

PC is a form of internalism about justification, which is the view that justification supervenes on the mental states of the subject, or things that are epistemically accessible. To say that justification supervenes onto the mental or the accessible is to say that there cannot be a change in mental states or what is accessible without a change in justificational status. The version of PC that takes the supervenience base to be mental states without an accessibility requirement is called mentalism, and it can be seen as a form of reductionism about justification. The version of PC that takes the base to be epistemically accessible things is called accessibilism, and is a version of non-reductionism about justification. Mentalism can give a reductive analysis of justification in terms of properties of mental states, whereas accessibilism takes access to be a primitive, epistemic notion which cannot be reductively analyzed without circularity. PC can be formulated in either way, but I take it to be a hybrid because seemings are both mental states and intrinsically accessible to the subject.

PC can be construed as either weak or strong foundationalism. If it is taken to be a version of weak foundationalism, then seemings are not sufficient for fully justified beliefs based on them. Beliefs based on seemings, on this view, would have some justification, but not enough for full blown justification. Those beliefs must also be supported by other beliefs, or other epistemically relevant states. If PC is a version of strong foundationalism, then seemings are sufficient for fully justified beliefs. Beliefs based on seemings are fully justified, absent defeaters. Huemer’s version of PC can be seen as a hybrid, where some seemings may not be sufficient for full justification, while others are. The hybrid nature of Huemer’s version of PC can be seen in the, “at least some degree of justification” clause.

Justified beliefs can be defeated by various considerations. PC allows for defeat, which means that beliefs based on seemings can lose their fully justified status. For example, if I look at a pencil submerged in a glass of water, it seems to me that the pencil is bent. Lacking background knowledge about what happens when straight objects are submerged in water, I form the belief that the pencil is bent. I now have a belief that is at least partially justified. But then I pull the pencil out of the water and see that it is not actually straight. Puzzled, I search wikipedia for an explanation, and learn about what happens when pencils are submerged in water. My belief about the pencil being bent is now defeated by counter evidence.

In some future posts I will explore objections to PC, such as the problem of cognitive penetrability, the Sellarsian dilemma, and the problem of the speckled hen. I will also examine issues related to the nature of seemings, and whether seemings form a homogeneous class of mental states, or if there are distinct kinds of seemings. Finally, I will explore the connection between PC and ethical intuitionism.

Works Cited

Huemer, Michael. “Compassionate Phenomenal Conservatism.” Philosophy and Phenomenological Research 74.1 (2007): 30-55. Web.


An Argument Against Moral Intuitionism

Moral intuitionism is usually characterized as the thesis that we have non-inferential moral knowledge. Any epistemological theory that posits non-inferential knowledge is a form of foundationalism, so moral intuitionism is a form of foundationalism. The thesis is usually accompanied by a description of the faculty of moral intuition. Sometimes moral intuition is considered a faculty of judgment that produces non-inferentially justified/warranted moral beliefs, some of which are true. Another characterization of moral intuition is as a special faculty of moral perception, analogous to vision (how far that analogy can be pushed depends on who you ask). All of these ways of describing moral intuition have their respective strengths and weaknesses, which is a topic for another time. In this post, I’m going to present an argument against moral intuitionism that does not assume any robust account of the faculty of moral intuition. All that is assumed for the sake of this argument regarding intuitionism is that it is a form of foundationalism; whether it’s internalist or externalist is immaterial to the thrust of the argument.

Now for the argument: There appear to be good reasons to think that our moral beliefs are formed under less-than epistemically appropriate conditions. Moral belief formation is supposedly subject to various cognitive biases and emotional influences. These various psychological phenomena, compounded by facts such as massive moral disagreement among seemingly rational people present us with good reason to think that many of our moral beliefs are probably false, or at the very least, unjustified/unwarranted.

Assuming that there is good evidence of these cognitive biases and emotional influences coming out of psychology and cognitive science, the moral intuitionist is presented with a dilemma. She is presented with a defeater for her moral beliefs in the form of evidence of the unreliable conditions under which they are formed. Either she can defeat this defeater or she cannot. If she cannot defeat the defeater, then she does not have non-inferential moral knowledge, which means some form of moral skepticism is true. If she attempts to defeat the defeater, then she must provide good reasons to think that her moral beliefs are formed under conditions conducive to their reliability. Let’s say she succeeds at defeating the defeater, and has given good reasons to think her moral beliefs are reliably formed. She has now provided a justificatory basis for her moral beliefs that renders her moral justification inferential. So, she does not have non-inferentially justified/warranted moral beliefs, but rather her moral beliefs are inferentially justified. So, either moral skepticism is the case or moral justification is inferential.

The intuitionist appears to be backed into a corner. However, things aren’t as they seem; she has two ways to resist the dilemma. The first way to avoid the conclusion is by challenging the principle that the defeater defeater must come in the form of evidence of the reliability of moral belief formation. Perhaps the defeater defeater could be evidence that moral belief formation is not influenced by the cognitive biases and emotional influences mentioned above. Note that that evidence is not evidence for the reliability of moral belief formation, but merely evidence against the case for their unreliability; so, the intuitionist using this strategy isn’t committed to the no non-inferential moral knowledge horn of the dilemma.

The second way to avoid the dilemma is by allowing for epistemically overdetermined beliefs; such beliefs gain justification/warrant from non-inferential and inferential sources. The intuitionist can allow for a defeater defeater that generates inferentially justified/warranted moral beliefs, while also claiming that such a defeater defeater restored non-inferential justification/warrant as well.

One may wonder what part the internalism/externalism distinction plays in this discussion. The argument appears to be neutral about whether or not justification/warrant/knowledge is extended. Even if some sort of reliabilism is true, the alleged evidence from psychology and cognitive science presents a potential defeater to one’s moral beliefs. So, it really doesn’t matter if one adopts internalism or externalism about moral knowledge.


Further Reading:

For some of the alleged evidence against moral intuitionism and various formulations of the argument presented above, see Walter Sinnott-Armstrong’s book Moral Skepticisms, and his papers, An Empirical Challenge to Moral Intuitionism, Framing Moral Intuitions, and Moral Intuitionism Meets Moral Psychology.

For a more developed response to Armstrong’s argument along the lines of the critiques I explored above, see Moral Intuitionism Defeated? by Nathan Ballantyne and Joshua Thurow.