Factivity and Deontic Modals

We have explored so far, in the previous posts, the differences between conceptual, epistemic, metaphysical and natural modalities, arguing that the two first have representational targets, but differ by the mind-dependence of their source, and that the two last have worldly targets. An important common characteristic of all these modalities is factivity: if P is necessary, in any sense of the term, then P is the case. The converse of this theorem of alethic modal logic is that if P is the case, then P is also possible, that is, compatible with the source of necessity. This is notably not the case of deontic and practical modalities: maybe it must be the case that no one walks on the grass according to the norms, but someone is doing it right now, or maybe we must stop the water from flowing into our basement, but we are not actually doing it. Deontic laws, contrarily to natural ones, can be broken.

Another way of making the difference is in terms of direction of fit. If someone walks on the grass, then the world must be changed to comply the law, not the other way around, but if an object starts falling towards the sky instead of falling towards the ground, nothing incorrect is happening. We must change what we think the law of nature to be, not what the object is doing. If a modality is factive, then witnessing a contradiction to the law is impossible, and so, we must adapt our representation to the world. If a modality is not factive, then the world can be adapted instead. In the case of epistemic modality, this translates into the incoherence of sentences such as "it is not raining, but (for all I know) it must be raining" (while "the door is not closed, but (according to the rules) it must be closed" makes perfect sense).

I would say, in line with the analysis of the distinction between epistemic and conceptual necessity from the previous post in terms of the mind-dependence of the source of necessity, that the right way of thinking of deontic modals is in terms of the source of necessity being representational or mind-dependent, associated with our aims and desires in particular, but their target being worldly. These are constraints from us onto reality (hence the direction of fit, opposite to the epistemic one), and since we are not omnipotent, such modality is not factive. This follows at least if we assimilate the deontic realm with our aims rather than with an external objective good. Practical modality is just the special case where deontic norms are construed more broadly as including our capacities (following an “ought implies can” reasoning) and local aims, thus restricting further the range of possibilities.

This means that conceptual/logical and practical/deontic necessity have something in common: their source lies within us, they are somehow normative, the only difference being that the former only constrains directly our thinking and reasoning, but the latter aims at constraining the world (this is in line with conceptual engineering ways of thinking, and opens the way to an interpretation that, I suspect, many would find unattractive: the idea that logical and conceptual modalities are non-factive, in the sense that we can fail to enforce their "law". However, this idea is not imposed upon us. I will address it in a future post, since it is related to a puzzle about conceptual change that was mentioned in the previous post). And deontic necessity also has something in common with natural one, which is that it affects worldly objects, hence the common reference to "laws".

We are now armed with two distinctions in order to classify modalities: (1) is the target of constraint representational or worldly? and (2) is the source of constraint representational or worldy? Accordingly, we get four kinds of modalities, which would be conceptual, epistemic, practical/deontic and natural/metaphysical. What we do not get with this way of thinking is a finer distinction between metaphysical and natural, or between practical and deontic modalities, and I will leave this issue aside in this work.

Source / Target	Representational	Worldly (law)
Mind (norm)	Logico-conceptual	Deontico-practical
World	Epistemic	Metaphysico-natural

What I will now be interested in is how this taxonomy interacts with modeling activities in science. Before moving to this theme, I will address in the next post a puzzle that occurred concerning conceptual possibilities and their difference from both deontic and epistemic ones.