Notes

Chapter 11. My paper on self-evidencing (Hohwy, 2016) appeared online in Noûs in 2014.In this book, there will be many references to the predictive pro cessing lit er a ture, with a slant toward philosophical treatments.My own book (Hohwy, 2013) and Andy Clark's influential paper and subsequent book (2013b, 2016) can be useful starting points.By 2020, a large body of philosophically oriented work on predictive processing had already emerged, with several hundred listed in the appendix of a review that I conducted (Hohwy, 2020a), and it has only grown since.Karl Friston articulated the free energy princi ple and is the architect behind its main developments; see, for example, his extremely influential landmark paper (Friston, 2010) and many other works by him, many of which I cite in this book.2. The free energy princi ple is a variational princi ple of least action; the recent body of technical lit er a ture thus makes it clear that it implies that there is an action for every existing system-namely, the path integral of variational free energy (Friston et al., 2023).For a discussion of the "high-road" versus "low-road" approaches to predictive pro cessing, see Friston (2019a); for a discussion of princi ples and pro cesses and references to the philosophy of science debates about the relation between princi ples and laws, see Hohwy (2021b), and for a critical discussion of the distinction, see Williams ( 2022).The metaphor of the high and low roads picks up on the idea that both will end up at the same destination, but via dif fer ent methodologies.Which methodology one prefers depends on pragmatic interests and scientific choices; the metaphor as it is employed in the lit er a ture on the free energy princi ple is not intended to say that either is better than the other.3. For many of the more philosophical critiques of predictive pro cessing and the free energy princi ple, there are already several published responses, which, as I see it, efficiently engage with the criticisms.See, for example, the commentaries by Friston, Ramstead, Andrews, Parr, Kiefer, and myself, as well as works by Seth, Clark, Kirchhoff, and several others, which were published with the following critiques of predictive pro cessing (Bruineberg et al.