Re: [R] [Obo-relations] Discussion summary on "original" biological parts

From: Wacek Kusnierczyk <Waclaw.Marcin.Kusnierczyk_at_idi.ntnu.no>
Date: Thu, 20 Nov 2008 17:56:14 +0100

mejino_at_u.washington.edu wrote:
>
>
>>> A canonical human body will have canonical parts and those canonical
>>> parts will
>>> have canonical subparts and so on.
>>
>> ... and?
>>
>>> Can't think of anyone who would fit that
>>> description.
>>
>> is this considered an argument for that there cannot possibly an
>> instance of the canonical human body?
>
> So what exactly is your understanding of a canonical entity and
> perhaps that'd clarify your point?

when i studied medicine, i used to think about 'canonical' anatomy (well, we'd just speak of human anatomy, with no canonicity referred to) as a model that corresponds to our expectations wrt. a human's body when we do not have any additional information. that is, the book descriptions reflect, in most cases, the most frequently occurring structural variants. there does not have to be a single human that is exactly as the book descriptions want (and there is *no* single canon in this respect, besides perhaps that a human has one heart and such gross stuff), but there is nothing in the way of there being a human that is accurately described by a particular canonical anatomy. that's why i got shot by your "There is no instance of a canonical human body or a canonical heart.", which does not make sense to me if it amounts to saying "there cannot be".

as to "A canonical human body will have canonical parts and those canonical parts will have canonical subparts and so on", the problem with the insisted canonicity here is that the more you go into details, the less canonicity to be found. if hardly anyone has more than one heart (and who has none, except for patients under surgery?), hardly any two people will have the same pattern of capillary vessels in a particular location in their body. that is, on the frequentist reading of 'canonicity', the gross level canonical descriptions correspond to strong accuracy of expectations, the more detailed levels correspond to weaker accuracy of expectations.

i have read the article on canonicity writen by fabian et al., and besides its logical clarity, i found the definitions completely useless. it might be that i read them too quickly, and had no time to examine them more carefully, but it stroke me that 'canonicity' was defeined there in complete dissociation from the frequentist view. basically, a canonical entity is one that is canonical.

>
> But let's assume the possibility that 1 or 2 out of 6 billion people
> fit the "idealized, canonical" type. I don't see much utility in that.

of course. what is useful is that if the canonical anatomy says that there is one heart in a human body, and that it is located here or there, then when a patients comes to me, without additional evidence i assume he/she has one heart and it is here or there. but i have less confidence when it comes to more detailed descriptions; these are more exemplary than canonical.

> I want a system that can accommodate and provide the information that
> your heart, my heart and anyone's heart are instances of some type
> "Heart" which defines the necessary and sufficient conditions that
> establish the identity of the heart structure in you, me and everyone
> else.

wait. if you define necessary conditions for an entity to be heart and call this canonical anatomy, then either there are no hearts (because none fulfils the necessary conditions) or there are hearts that instantiate canonical anatomy. i must be wrong, but where.

> This is why I'm re-evaluating my position with regards to any
> canonical reference to the FMA, unless of course we redefine
> "canonicity" to mean some general or generic description.

my feeling is that the term 'canonical' is virtually meaningless the way you seem to use it.

vQ



R-help_at_r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. Received on Thu 20 Nov 2008 - 17:00:29 GMT

Archive maintained by Robert King, hosted by the discipline of statistics at the University of Newcastle, Australia.
Archive generated by hypermail 2.2.0, at Thu 20 Nov 2008 - 20:30:30 GMT.

Mailing list information is available at https://stat.ethz.ch/mailman/listinfo/r-help. Please read the posting guide before posting to the list.

list of date sections of archive