Wednesday, October 26, 2011

On not multipying entities

There is a nice post today by the Maverick on “The Use and Abuse of Occam's Razor: On Multiplying Entities Beyond Necessity” There are few points to raise. Maverick writes “Occam's Razor is standardly taken to be a principle of theoretical economy or parsimony that states: Do not multiply entities beyond necessity.” True, it is standardly taken thus, but as Thorburn showed nearly 100 years ago, Ockham did not say exactly that. He actually said that plurality is not to be posited without necessity (Pluralitas non est ponenda sine necessitate). Moreover, it is not ‘his’ razor. Scotus (on the lines of whose thinking Ockham’s thinking is largely developed) used it, and it is probably earlier than that. He also says that is vain to bring about through more what can be brought about by fewer (frustra fit per plura quod potest fieri per pauciora).

Furthermore, the maxim does not really capture the spirit of Ockham’s nominalism, which is better expressed by his claim that one cause of error is ‘to multiply entities according to the multiplicity of terms, and that every term has a (corresponding) real essence’ (Secunda radix est multiplicare entia secundum multitudinem terminorum, et quod quilibet terminus habet quid rei).

He says this at the end of chapter 51 of the monumental and magnificient Summa Logicae, of whose structure you can get a flavour here. Chapters 40-62 are a long discussion of Aristotle’s categories, and Ockham’s objective, after some essential preliminaries set out in chapters 1-17, is to show that most of the ten categories are not really types of being at all, but really types of term. For example, chapter 51 is part of chapters 49-54 on Aristotle’s category of relation (ad aliquid, relatio). Ockham wants to show that the term ‘relation’ is not a name for a particular type of thing, outside the mind, really distinct from some absolute thing (res extra animam, distincta realiter a re absoluta). Otherwise, whenever a donkey moved down on earth below, every heavenly body would be changed in itself, because of the change in its spatial relation with the donkey. Or we might mistakenly suppose that a father is a father by some extramental thing such as ‘paternity’.

We are led into these errors from the ease with which Latin (and other romance languages, but Ockham rarely talks about these) is able to construct abstract terms like ‘fatherhood’ from concrete terms like ‘father’. He discusses this in chapter 5 and subsequently. Such terms have a similar beginning verbally, but different endings, and the abstract nearly always has more syllables than the concrete. Ockham argues that the concrete and the abstract are really synonyms. To say that Socrates has humanity is no more than to say that Socrates is a man. For this reason there are no abstract names corresponding to many concrete names. E.g. though we frequently use the names ‘cow’, ‘donkey’, ‘goat’, there are no corresponding abstract terms like ‘cowhood’ or ‘donkeyness’. And the ancient philosophers did not use this diversity “except as an ornament of speech, or for some other accidental reason, just as in the case of synonymous names.” [--]

No comments: