From: Miguel Carrasquer
Message: 39845
Date: 2005-09-01
>--- Miguel Carrasquer <mcv@...> wrote:So it *is* a single rule.
>
>> On Tue, 30 Aug 2005 06:52:31 -0700 (PDT), Sean
>> Whalen
>> <stlatos@...> wrote:
>
>> >--- Miguel Carrasquer <mcv@...> wrote:
>> >
>> >> On Mon, 29 Aug 2005 02:20:29 -0700 (PDT), Sean
>> >> Whalen
>> >> <stlatos@...> wrote:
>> >
>> >> >C-sonorant-continuant > -aspirate / _ C-aspirate
>> 2
>> >>
>> >> What is the use of this one?
>> >
>> > It doesn't change anything for this set of rules
>> >considered all at once but it's likely a long time
>> >passed between this and the next relevant rule.
>>
>> That doesn't make it less unnecessary.
>>
>> > At the stage it takes place there would be a
>> >distinction between voiceless, voiced, and voiced
>> >aspirated. The aspirates would devoice before
>> >voiceless
>>
>> Creating a fourth category?
>
> No, in my theory these rules operate immediately
>one after the other.
>> >and then deaspirate (bhranghtas>bhrankhtas>You say yourself that there *is* no intermediate stage.
>> >bhranktas) so the distinctions would become erased
>> >before the unmarked series.
>>
>> In a system without voiceless aspirates (without
>> distinction
>> between voiceless aspirates and non-aspirates), /gh/
>> would
>> have been devoiced to /k/ by a single rule.
>
> After reading this I don't understand what you were
>originally criticizing. Do you mean I should have
>used one rule like:
>
>C > -voice-aspirate / _ C-voice (-aspirate unmarked)
>
> I think that in describing a reconstruction the
>rules should be separated as much as possible for ease
>of understanding, to show the intermediate stages
>which might be needed in a synchronic analysis, and inIf you put rule 3 first, you don't need rule 2.
>case more evidence causes the need for a change in
>one. They can always be combined if necessary to
>provide evidence for a theory, etc.
>
> There's also the possibility that the voiceless
>aspirates that were allophones of voiceless stops in
>PIE survived into this stage of the proto-language.
>In usual reconstructions a rule like:
>
>C-voice > -aspirate
>
>would be included since they have the same outcome in
>Germanic as the plain voiceless. However, in my
>theory voiceless stops become aspirated later so it's
>possible the original voiceless stops were retained up
>to this point. I have no way of discovering this
>through analysis of the Germanic outcomes so I made
>rules that work either way.
> I also don't know what time period these changesOn the other hand, rule 1 already excludes Indo-Iranian,
>are from. If the earlier rules occurred in dialects
>that would become Proto-Germanic and -Italic, for
>example, both rules _would_ be necessary since Italic
>didn't aspirate the plain voicelss.