Monday, 15 February 2016

Emrys Westacott's Existential fallacy

What kind of fallacy is Prof.Emrys Westacott guilty of here?
'...we have separated out several different kinds of respect. These are: (1) respect for a person in virtue of what they share with all humanity (2) respect for a person in virtue of their specific qualities, achievements, or experiences (3) respect for a person's right to hold a belief (4) respect for a particular belief A lot of people assume that both (1) and (3) entail (4); so they worry that to withhold (4) implies that one is withholding at least (1) and (3). This is perhaps what underlies much ready talk about respecting beliefs. The worry is perhaps exacerbated by the fact that withholding (4) may well involve a diminishment of (2).'

Since Respect has to do with a prudential regard and heedfulness, it is linked to 'regret minimization' as opposed to 'utility maximization'.  We may respect the corporeal inviolability of the golden goose, by not slitting its gizzard, because we will later regret losing its golden eggs. We may respect the numinous sanctity of the Capitoline geese because, it may be, they can sense things we can not- perhaps an invasion, but perhaps also an earthquake. Respect for the Environment or for animals or for oracles or autistic savants may be prudential and 'regret minimizing'.

A special instance of Respect arises in the case of autonomous agents capable of rational information processing and intelligible discourse.
Respect for such persons as 'ends in themselves' may well be 'regret minimizing' even though, in the short run, some benefit is not ruthlessly extracted from them.
Such respect may even be something 'shared with all humanity' in the special sense that there is a reflective equilibrium all reasonable people would agree to and thus we can assume a 'Revelation Principle' such that strategic problems need not detain or deter us from eusocial Mechanism Design or the deontic or doxastic reasoning in which such Mechanism Design is embedded or expressed.

The problem here is that if we knew what a given person shares 'with all humanity' then we would have a metric for 'respect' (i.e. the vector associated with the relevant reflective equilibrium's Hannan Consistent information aggregating multiplicative weights update algorithm) which, sooner or later, all people with equal knowledge would concur with. There would be 'Aumann Agreement' as to the extensional content, both general and specific, of the word 'respect'. It would be well defined and eminently operationalizable.

Clearly no such knowledge actually obtains. Indeed, it can not obtain if Knightian Uncertainty exists and human beings evolved by Natural Section. Thus (1) is vacuous. It commits the existential fallacy. For a given human, what is shared with all Humanity, and which is also of salience in this context, might well be the empty set.

 If human beings really did evolve from non-humans, who in turn rose up ultimately from inorganic matter, then the underlying set is impredicative. It may also, for all we know, be infinite. Thus, there is no warrant for assuming away the existential fallacy for the purpose of a elaborating a deontological comparative statics.

Still, purely as a doxastic exercise, let us grant (1)- with the result that having a non empty doxastic endowment is assumed to be a property of every agent. 

 The problem here is that (2)- i.e. discriminating a specific person's 'qualities, achievements and experiences'- requires us to ascribe an epistemic content to (1). Indeed, we might say there is a dialectical relationship between (1) and (2). The more we learn about the specific individual's haecceity, the more our idea of 'what is shared with all humanity' changes and this is a continuous and reciprocal process as more individuals fall within our ken.

One way out is to end this doxastic exercise immediately with the acknowledgment that this sort of thing is better covered by Game Theory- more particularly because strategic considerations arise under the rubric of Newcombe's problem, Kakfa's toxin and so on. If there is any Evolutionary benefit from 'doxastic' dialectic, it can be illuminated by the Mathematical and Quantitative Sciences. We can turn back to Parmenides and let doxa make its own way in the world.

However, it appears that Prof Westacott knows of people who believe (3)- i.e. that a duty to respect another person's right to subscribe to a doxastic system exists. Surely this duty could be discharged most efficaciously by abstaining from evangelizing for one's own favored belief system? Don't ask, don't tell- least said, soonest mended. In this case, no obnoxious dilemma related to (4)- viz. respecting particular beliefs- need arise.


  1. Utterly wrong. There is no 'existential fallacy' in what you quote since it is not a logical proposition but part of a specific i-Language Game. Kindly look up the Stanford article by the Professor-
    It wouldn't matter in the least if 'For a given human, what is shared with all Humanity, and which is also of salience in this context, might well be the empty set.' because the subject is at liberty to introduce any convention she pleases to guarantee the reverse. Notice, 'Respect' is defined as something generated autonomously by the subject. Categorical arguments, in this context, can at best generate antinomies. They can't be shown to be fallacious simply.

  2. The point about i-language games is that they are notoriously open AND defeasible. Otherwise they wouldn't be games but substantive simply.
    There is no question that the learned Professor is committing to an existential fallacy. This is because there is an alethic aspect to his argument.
    I have read the article you link to and consider it risible. So does everybody else- in practice.