I have worked on and developed several epistemological and methodological views and theses during the last few years. Not all of them are prima facie compatible, however. Here, I want to show that, in fact, they are, and thus complete some parts of the picture. I do this by taking on some apparent inconsistencies one by one, in (1), (2) and (3), before concluding in (4). I write all this in public because different people have read or heard different things that I have written or said, so it would be nice to be able to point to this post to clear up possible misunderstandings, and, moreover, readers who are new to the picture, I may add, may learn something about its main content while I attempt to do so.

(1)

The first apparent inconsistency stems from accepting epistemic holism – as a part of an account of epistemic justification – and also wanting to claim that, in fact, metaethics (or second-order ethics) is epistemologically prior to normative ethics (or first-order ethics). For, prima facie, on a holist view, all kinds of views are relevant for the justification of all others. But then it seems that first-order normative beliefs also could partake in justifying (second-order) descriptive beliefs (about those normative beliefs, or in general). There are even authors who explicitly argue in this way.[1]

Now, on my (still holistic) account of justification, I do not claim that all claims count for equal in justification, as many defenders of reflective equilibrium or coherentism as theories of justification or metaphilosophies do. I hold that, in so far as they are parts of people’s sets of webs-of-evidential- states (rather than just web-of-beliefs), some kinds of intuitions – Huemer-Bealer intuitions,[2] as well as a kind of normative intuitions I call Insights – and some kinds of perceptions count for much more in justification than other such states or mere speculative theory.

Unfortunately, that may not appear to be neither here nor there, because said kinds of intuitions may well be held to be the relevant kind of moral intuitions. As a matter of fact, I reject that idea. I do not think that the relevant, substantive kinds of moral intuitions exist (nor do moral properties). But those are substantive metaethical theses, and it seems question-begging of me to argue from metaethics to a primacy of metaethics.

So let us lift our gazes to metametaethics. Quite generally, I think metaphilosophy is to philosophize about some aspect of philosophy (where that aspect could also be disjunctive or conjunctive or all of the field). As ethics is part of philosophy, it may be thought that metaethics is then all the relevant metaphilosophy there could be to ethics. But metaethics is not just that. It is to think about second-order questions about first-order ethics while using (inter alia, though primarily) the tools of the LEMMings subjects. And how to do that can also be metaphilosophized about. And I think there are interesting metametaethical questions to ask without requiring substantive commitments to theses in either metaethics or ordinary ethics.

Once here, then, I think, it is not just coherent but very plausible to hold that first- and second order ethics in fact do have impact on each others, but that the impact of second-order ethics on first-order ethics ought to count as much greater than the other way around, not least due to its greater abstraction and thus ability to be able to touch more issues in first-order ethics than a purely first-order ethics could do. This is for the reasons just mentioned, but also due to the arguments I provide here. So I have not denied, then, that first-order and second-order ethics both can have impacts on each other. But as a higher-order thought about both fields at once, it seems that the impact of second-order ethics can be much greater on first-order ethics than the other way around. And that rescues my arguments.

(2)

The second apparent inconsistency stems from holding, first, as mentioned in (1), that intuitions and perception counts for more than theory in the amount of justification they grant for a view, and, second, claiming that, in fact, moral thinking ought to be theory-driven. How could this be?

There are several points to make to dissolve this apparent dilemma. But I need to qualify my view here to allow for complications. I have been vague and not made everything out clearly; I should have written that intuitions and perception count for more, ceteris paribus. There can certainly be situations where they do not count for more. For example, a person having an hallucination is justified in believing that her perception is a hallucination in virtue of knowing how hallucinations work in theory. So, pending defeaters for the ordinarily more important sources of evidence or their contraries for the usually lower counting ones, the view stands. For it is still certainly true that in the ordinary case, an empirically well-supported theory is better than one that is not, and that I am justified in believing a purported proof that I can grasp intuitively but less justified in believing one I cannot grasp (and no experts tell me trust, etc).

Now, when I wrote about the theoretical virtues of theory-driven ethics, I wrote, specifically, “first, metaethical implications in normative theory count for a lot, and second, in so far as one does  the normative stuff, one should work abstractly, preferring going from principles to cases, rather than the other way around.” By contrast, when I think about theory-driven thinking elsewhere, I think about undersupported theoretical ideas or hypotheses – astrological thoughts, Myers-Briggs psychology, naïve appeals to rational choice, and similar more or less pseudo-scientific ideas. It is clear that this kind of theory, in the latter sense, does not do much to justify any beliefs about, e.g., their own implications for practical living.

The difference, then, seems to be one between justified and unjustified appeals to theory, or, where things are or are not equal. A theory which is strongly supported by good first-order evidence or arguments is often a good argument in some cases: here, all else is not equal. But it would not have been so if it did not have good independent, first-order support. But this the pseudo-scientific ideas decried here certainly lack, whereas metaethics as a field certainly stands a good deal stronger than that of first-order ethics – for the reasons mentioned in (1) – and thus seems justified to appeal to. This accounts for metaethically informed normative theory. That accounts for the first point about theory in practical normative theorization.

On the second point, however, one may prima facie believe that I cannot make sense of what I said, for a theory-driven approach in normative ethics is certainly one which seems to against the grain of the epistemology I defended. Here, however, I also argued: “I think the practice-driven views are non-starters: to specify what is good or bad, right or wrong, just or unjust in particular situations you need the general theories in the first place. And, furthermore, Ross-style intuitionist pluralisms tend to strike me as cheating: why care about, not to mention how adjudicate between, particular intuitions unless you have some general systematic theory to fall back on? Surely a more consistent theory is preferable to another, and if it is based on some general value, it can give some support about what to do in many cases of adjudication between various principles.”  Thus, I appealed to intuitions to undercut some other intuitions. And that is, of course, compatible with my general view. Theory-driven approaches to cases are justified in virtue of general intuitive considerations – or so they would be, that is, for moral realists or others who can argue in such situations. I am not one; I believe distinctively moral intuitions are strong knee-jerk gut reactions (as per social intuitionism) and that there are no moral properties, despite moral language attempting to be descriptive. So I was arguing hypothetically about what they should do on their own premises rather than from my own.[3]

(3)

The third and final apparent inconsistency is this: in (meta-)normative contexts, in particular in epistemology, my method of choice is often fairly interpretive. But in a recent text[4] I have also defended social scientific methodological positivism (or, more specifically, I have done so in the context of writing about philosophy of law, but my argument – primarily contra Ronald Dworkin’s quietist, internal-perspective realism – generalizes to, inter alia, metaethics as well). How could this be? Is there not a trade-off here: either I treat knowledge as a (natural? social?) kind just as I treat law as a social kind, or I treat both law and epistemology as to be interpreted and (thus) evaluated?

My answer here is that law and knowledge are distinct phenomena. It is easy, when describing various social systems, to imagine a society (or pattern of interactions more broadly, such as the international system) where there are no laws whatsoever – an anarchy. Thus, law qua social institution is clearly to some extent dependent on convention.[5] By contrast, when doing epistemology one is necessarily trying to contribute to human knowledge – in this case, human knowledge about (the phenomenon of) human knowledge, since one aims at getting to know something about the nature of knowledge. This suggests that, to the extent that knowledge is a kind (and irrespectively of which kind of a kind it is), it is one in which the epistemologist also is bound up.

Dworkin has argued that legal philosophers, inquiring into the nature of law are in the position of judges in so far as they argue about what law (“in the doctrinal sense”) is. To do so, they have to bring in their (moral) values, and law is ultimately constituted by the best moral interpretation of currently existing legal texts.[6] I have argued that this is not the way to proceed, however, since, in fact, in so far as we qua inquirers try to interpret the world, we are better off trying to interpret its phenomena as posited in the world and then quantify over them, in the usual scientific manner. Thus, even in so far as one aims to interpret the social phenomenon of law, one needs take the active role of a judge. And in so far as interpretation of law and the behaviour of jurists is needed to understand law in this kind of way – which it, indeed, often is – the inquirer needs not be a full participant in the practice of law-making. For even though an inquirer takes up one kind of internal or first-person perspective to what she is doing in so far as she aims to reach the best understanding of what is going on, this kind of internal perspective needs not commit one to the internal perspective of lawyers in evaluating the law, rather than partially approximating it to see how they evaluate it.

Here knowledge differs. Any inquirer into the phenomenon of knowledge is committed to trying to gain some kind of knowledge, since that is just what it is to do (rational) inquiry. Thus, the internal perspective of epistemologists is a starting point for inquiry into knowledge that one needs to presuppose is not arbitrary, on pain of being unable to carry out the kind of inquiry one is doing. So while an interpreter of law needs not think she is a jurist, an interpreter of knowledge needs to think she is a knower.

But I said “needs to think she is”. For as the possibility of error theories show, it is not pre-theoretically obvious that necessary commitments are right: our commitments can, of course, be systematically false. My reply is then that here, an epistemologist must try to integrate her internal perspective on the practice In which she is partaking with the rest of the world. Here a plethora of metanormative views open themselves up: everything that metaethicists do, metaepistemologists can do as well. My preferred approach here is to try to reconcile, on the one hand, practical and theoretical normativity, and on the other, the normative and descriptive perspectives that can be taken by a kind of naturalistic reduction of normativity to not reasons but rationality. To this extent, I am a “positivist” in a broad sense here as well as I aim not only to do first-order epistemology but also to reach an understanding of knowledge in the world – although a positivist whose starting point is interpretation of the phenomena.

And precisely that was also the methodological stance presupposed in the social scientific positivism I defended, although, after starting to interpret the phenomena, a kind of reconciliation with the natural world is necessary in both cases, but that kind of reconciliation yields – I think – a debunking understanding of legal truth-makers in so far as law turns out to be social phenomena of little intrinsic normative force when interpreted and then “operationalized”, but a vindictive one of epistemic truth-makers in so far as that it turns out that the normative facts indeed exist as natural facts.

(4)

In (1), I argued that there was no inconsistency between preferring second-order inquiry into ethics  above first-order inquiry because my arguments were meta-metaethical rather than metaethical. In (2), there was no inconsistency because I should have qualified my view to allow for defeaters (and “winners”) and that what else, in fact, seemed inconsistent actually was consistent due to the intuitive argumentation underlying the theories. And in (3), I argued that there was no inconsistency between being fairly interpretive in epistemology and fairly positivistic in the philosophy of law, because the phenomena were different and that, more abstractly formulated, the starting point of inquiry into both is interpretive, but, upon further interpretation, it turns out that legal norms ought to be given a debunking explanation as social facts, whereas epistemic facts can be given a vindicating explanation.

So far, so good, then. I shall now lay the epistemological and methodological issues to rest for quite some time, focusing instead on more substantive – primarily metaphysical, in some broad sense of that word – issues in metaethics and philosophy of action. For now, enough of epistemology.



[1] For example, Enoch, David (2011), Taking Morality Seriously, Oxford, UK: Oxford University Press.

[2] See e.g. Huemer, Michael (2005), Ethical Intuitionism, New York, NY: Palgrave MacMillan, and Bealer, George (1998). “Intuition and the Autonomy of Philosophy”, in DePaul, Michael & Ramsey, William (eds.) (1998), Rethinking Intuition – The Psychology of Intuition and Its Role in Philosophical Inquiry, Lanham, ML: Rowman & Littlefield., pp. 201-239.

[3] My own (re-)constructivist view is that a ”moral” theory ought to make sense of the Insights mentioned above, but that these do not stem from arguing about cases or aim to represent the world (but fail, because it hits nothing), as case-driven moral intuiting does. And that is, again, compatible with my epistemic view.

[4] About the Hart-Dworkin debate – it is available upon request.

[5] I say ”qua social institution” so as not to beg the question against natural law theorists. But I doubt that I would do so anyway even if I said this about law more broadly: sophisticated natural law theories rather say that the function of legal systems is moral than that there is a natural law that they necessarily instantiate.

[6] See, e.g., Dworkin, Ronald (1986), Law’s Empire, Oxford, UK: Hart Publishing.