[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Subject Index][Author Index]

How truncation works, and how ranks don't work was Re: Sauropodz r kewl



Am 18.04.2012 15:53, schrieb Robert Schenck:

 On Wed, Apr 18, 2012 at 9:33 AM, Anthony Docimo <keenir@hotmail.com>
 wrote:
>
> hoping the paragraph breaks don't vanish again.
>
 I believe this is because you sent an email to the list as in a
 'rich text', where it actually only accepts plain text. Rich text
 messages show up on the list as just a message saying 'this was rich
 text and can't be shown'. Somehow D.M. was able to retrieve it to
 respond.

Anthony sent the e-mail in both plain text _and_ some other format. These are two successive parts of the same e-mail ("a multipart MIME message"). The filter of the list cuts the part in the other format away, replaces it with the "truncated" message, and then passes the e-mail on; it still has two parts. Some webmail interfaces and e-mail programs are sensible enough to show the plain text; others (like Thunderbird), however, follow the standard, which says to display the last part of the e-mail -- the truncation message.

What I do is press Ctrl+U, which shows the source text of the e-mail in a new window. This displays both parts and the header. I copy & paste the plain-text part into a new e-mail and send it as plain text only. Usually the plain-text part is derived from the other part in ways that represent special characters, even punctuation and line breaks, in very annoying ways; that's why it's so garbled most of the time.

Otherwise plain text without line breaks (no idea how that format is actually called) is treated as _not plain text_ by the filter. If that's what you get when you click on "reply", you must manually change the format to plain text (again).

Outlook Express formats replies like the original. Thunderbird, annoyingly, formats them all the same way (HTML or plain text), you can only choose which way; if I set that to plain text, I couldn't send HTML messages at all anymore, so I set it to HTML and manually change the format of everything I send to the list to "plain text only" (not, mind you, to "determine automatically", which would send it both ways and trigger truncation).

> all the extant organisms as closely related to the tuatara as mice
> are related to muskrats

 In a way we can do that, we can build a phylogenetic tree with
 quantified branch lengths, calculate the length from mice to
 muskrats, get a number, and then find all taxa at the same difference
 from the tuatara. Now, given two researchers, they'll both produce
 trees with different branch lengths and thus get different answers.
 Hell given two researchers you'll probably get two different tree
 topologies too!

And what do your branch lengths represent? The number of evolutionary changes? That depends (to varying degrees) on the dataset you used even if you used the same methods to calculate the tree. The age of the branch? There we get into all the trouble of divergence-dating methods.

> That doesn't sound like a good deal, does it? Because sometimes
> things *are* distinct and separate.

 I am sure D.M. will respond, but for my part, with living organisms,
 we have a species concept that allows us to quantify species; if
 they can't interbreed, they're distinct species. If they can
 fruitfully interbreed, they're just varieties of the same species.
 There is no way to quantify the difference, in a similar way,
 between genera or families.

There's not just one way to quantify species, there are lots of them*, and they all give different results; but, as you say, the other ranks have never been defined at all.

(...yeah... on very rare occasions, there have been proposals to extend the "Biological Species Concept"s to genera: species that can have viable -- even if infertile -- offspring with each other are in the same genus. I don't think this concept has _ever_ been used by anyone. That's why there are intergeneric hybrids in the literature.)

* ...even though not all 147 species concepts or however many there are today allow quantification. Many are vague and subjective.

> I don't need a yardstick to tell me the Atlantic is broader than
> the Mississipi and the Rhine

But you need a yardstick to tell whether the Atlantic is closer to 5000 or closer to 10000 yards wide.

You _can't_ use a yardstick or any other instrument to tell whether a taxon should be ranked as a genus, a family, an order or whatever. There is no genericometer. (Search the archives of this list for that word.)

 But of course you are making a qualitative determination, not a
 quantitative one. We can't quantify how much difference is required
 between two populations to determine if they are in different genera
 (and similarly we have trouble quantifying the the degree of
 difference for sub-species distinctions).

Not "we have trouble", but "the category of subspecies has been defined in several different ways, few of which are ever actually used, and some species concepts make subspecies impossible...".

>> Oh, there have been a few extremely short-lived attempts to
>> define ranks as times of origins of taxa: everything that
>> originated in the Cretaceous would become, I don't know, an
>> order, and so on.
>
> I believe the technical response is "WTF?"

 I think there's some sense to this (even if we ultimately reject
 it). Consider the idea that most phyla appeared in the Cambrian. If
 a population of lizards separated into two species today, you'd have
 a hard time convincing anyone that one of them was a new phyla. So
 the differences that make up the phyla represent 'huge' differences
 that accumulated over a long period of time, and therefore the phyla
 originated in the pre-Cambrian. Orders in the Cretaceous, etc. Again,
 not advocating that, but I think that's the idea behind it.

Yep, that was the idea. It would cause enormous upheaval if implemented, and it would almost always assign identical ranks to several nested taxa*, so it hasn't been implemented and won't be.

* Let's see... Avebrevicauda, Carinatae in the widest sense, Pygostylia, Ornithothoraces, Euornithes, Ornithuromorpha, Ornithurae in most senses including the strictest, and maybe Carinatae in the strictest sense all originated in the Early Cretaceous.