Rodlox wrote at 2004-10-15 00:13:05 (+0200)
 > ----- Original Message -----
 > From: Tim May <[log in to unmask]>
 > To: <[log in to unmask]>
 > Sent: Thursday, October 14, 2004 5:06 PM
 > Subject: Re: initial mutation or trigger? Re: re Mutations
 >
 >
 > > Rodlox wrote at 2004-10-14 17:40:36 (+0200)
 > >  > >
 > >  > > As I read it, "initial consonant mutation" isn't any kind of
 > >  > > technical linguistics terminology; it's just plain English.
 > >  > > The initial consonant (e.g. /p/) undergoes a mutation (to
 > >  > > /b/), therefore it's called initial consonant mutation.  I
 > >  > > suppose could have been called "first consonant change", but
 > >  > > I can't think of any way to make it plainer than that.  What
 > >  > > the heck does a phonetic change like that have to do with
 > >  > > triggers?
 > >  >
 > >  >  it sounds like the explanation I got (way earlier) regarding
 > >  > triggers...only it's attached to a word.
 > >  >
 > >
 > > Explain what you now understand a trigger to be.
 >
 >  an isolate either at the beginning of a sentance (common), or just
 > before the word...and the trigger can change the meaning of a
 > sentance by simply being one letter different.

Ok... this isn't right.  I'm not even sure what it means, but it's
certainly not a correct definition of "trigger".

I'm not sure what you mean by "isolate".  The only linguistic use
of that term I know is a language without an (known) relatives,
i.e. one that is not a member of any language family".  Clearly that's
not what you mean here.  And you don't really say anything about what
this thing does, except that it goes before a sentence, or before a
word... it seems like a very vague category.  So vague I'm not sure
whether either the Welsh mutation example or Tagalog triggers would
fall under it or not...

Anyway, I will now attempt to define "trigger".  Austronesianists and
Tagalog speakers, stand by to correct me.

"Trigger" is a term which is sometimes used in describing the grammar
of certain languages of the Austronesian family.  The terms "Philippine-type
language" and "symmetrical voice language" are more-or-less
synonymous with "trigger language", as I understand them, although
they may be used differently by different linguists.  In any case,
Tagalog (spoken in the Philippines) is one such language.

Anyway, although Tagalog has been fairly extensively studied by
linguists, there has been a great deal of disagreement on how best to
describe certain features of its grammar.[1]

So, bearing this in mind, what is a trigger?

Well, "trigger" is a category roughly analogous to that of "subject".
In a simple Tagalog sentence, you have a verb, and a number of nouns
phrases, preceded by particles.  One of these is preceded by the
particle "ang"[2].  This ang-phrase is the "trigger".

 | In previous linguistic studies of Tagalog, the ang-phrase has been
 | labeled topic, trigger, focus, subject, pivot, nominative, or
 | absolutive.
   [3]

The semantic role of the referent of the trigger in the sentence is
determined by a marking on the verb.  I believe Schachter, who I think
came up with the term "trigger", called this "focus" - most other
descriptions call it "voice".  (I'll use the terms interchangeably.)
So for example, if you want to say "I gave the box to the boy", you
could make "I" the trigger, and put the verb "give" in agent voice -
the trigger is the agent.  Or you could make "box" the trigger, and
put the verb in patient voice - the trigger is the patient.  Or you
could make "boy" the trigger, and put the verb in locative voice - the
trigger is the location (more of a recipient, here, but the same form
is used).  Here are those examples in Tagalog, as produced and glossed
by Kristian Jensen in an old Conlang post:

 | AGENT TRIGGER
 |      bumigay  ako   ng  cahon sa  bata
 |      give:AT  1:TRG GEN box   OBL child
 |      lit.: "I was the giver of the/a box to a/the child."
 |      "I gave the box to the child."
 |
 | PATIENT TRIGGER
 |      binigay  ko    ang kahon sa  bata
 |      give:PT  1:GEN TRG box   OBL child
 |      lit.: "The box was my 'give' to a/the child."
 |      "The box is such that I gave it to the child"
 |
 | LOCATION (DIRECTION OF ACTION) TRIGGER
 |      binigyan ko    ng  kahon ang bata
 |      give:LT  1:GEN GEN box   TRG child
 |      lit.: "The child was my 'giving-place/recipient' of a/the box."
 |      "The child is such that I gave him a box."

What motivates the choice of trigger?  The details aren't entirely
clear to me.  Clearly it's largely a matter of emphasis and
pragmatics.  The trigger is normally the pragmatic topic, I think, in
the sense Ray recently gave from Trask[5].  Also, rather
interestingly, the trigger is normally translated as definite:

 | Mag-kudkod=ka     ng=niyog.
 | AT-grate=2SG.TRG  GEN=coconut
 | ‘(You) grate some coconut.’

vs

 | Kudkur-in=mo     ang=niyog.
 | grate-PT=2SG.GEN TRG=coconut
 | ‘(You) grate the coconut.

   (Example taken from [6] - I've altered the glosses to resemble
   Kristian's above.  The ='s indicate that the pronouns and "ang" are
   actually clitics, but that's incidental.)

Now, of course, you can see that this all looks not so different from
the voice system of a familiar nominative/accusative language -  the
trigger is the subject, marked for nominative case; "agent focus" is
active voice; "patient focus" is passive voice.  Or, alternatively,
the system is ergative - the trigger is absolutive; "patient focus" is
the basic voice (not sure what you call that in an ergative system);
"agent focus" is antipassive voice.  And, indeed, Tagalog has been
analysed in both these ways.  But if it's a nominative or an ergative
system, it's rather different from those we're used to.

1. Neither agent nor patient voice is clearly more marked than the
   other, and both require an affix to be added to the root.  You
   expect passive or antipassive voice to be more basic than the
   alternative, in a nominative or ergative.

2. "Voice" changes don't change the valency of the verb.  If you put a
   verb in the passive voice in an accusative language, then the
   accusative argument becomes nominative and the erstwhile subject is
   no longer necessary.  If it is expressed, it's demoted to an
   oblique.  Similarly, the antipassive promotes an ergative to
   absolutive and demotes the absolutive.  But non-trigger core
   arguments are always marked with the genitive particle "ng" in
   Tagalog, and the valency of the verb does not change.

3. There are voices which allow oblique arguments like instruments and
   locations to be the trigger, without any additional derivational
   morphology.

4. The trigger doesn't pass all the normal tests for subjecthood - in
   some cases the agent (which may or may not be the trigger) acts in
   the manner expected of a subject.  I forget the details, though.

Well, that should be enough to get the gist of what "trigger" means,
in the context of Tagalog.  All this ignores the interesting issue of
whether these are really "nouns" and "verbs" at all, in the familiar
sense.  For further reading, see "Symmetrical voice systems and
Precategoriality in Philippine Languages"[7] (Foley), "Lexical
Categories and Voice in Tagalog"[8], "Tagalog"[9], and "The
Austronesian languages of Asia and Madagascar: Typological
characteristics"[10] (all Himmelmann).



[1] As Nikolaus Himmelmann puts it:

 | Among the factors which contribute to this lack of a common
 | grammaticographic practice is the following paradox: Philippine
 | languages are very similar to, and at the same time very different
 | from, Indo-European languages, on which all (western)
 | grammaticographic practices are based. When working on isolating
 | Chinese or polysynthetic Cayuga, the differences to Standard
 | Average European are almost immediately obvious and it is clear
 | that these require major adjustments of Greek and Latin-based
 | grammatical categories. This is not so in the case of Tagalog: in
 | some sense Tagalog has voice alternations, in another it has not;
 | there is evidence for the grammatical relation subject, the lexical
 | categories noun, verb and adjective, a distinction between
 | derivation and inflection, nominal case marking, core
 | vs. peripheral arguments, etc.; but there is, equally, lots of
 | counterevidence against these categories (see Himmelmann 2004 for
 | examples, discussion and references). Or, to put this in a perhaps
 | more productive way, most of the basic categories of Standard
 | Average European are also found in Tagalog, but in an interestingly
 | different way.
 |
 | Footnote: One consequence of this state of affairs is that in every
 | single case a decision is required of whether or not to apply the
 | grammatical terminology established in the description of Standard
 | Average European to a given phenomenon in Tagalog. The alternative
 | is the creation of new, often idiosyncratic terminology by
 | reinterpreting standard terms (e.g., focus instead of voice, topic
 | for subject) or creating new ones (e.g. trigger for subject,
 | transient word for verb). The present contribution is
 | terminologically conservative, opting in most cases for the use of
 | ‘traditional’ terminology.
 From
 http://www.linguistics.ruhr-uni-bochum.de/~himmelmann/multifunctional_ma_07-04.pdf


[2] If it's a common noun.  Proper nouns and pronouns are marked in an
    equivalent but different fashion.

[3] http://www.nevada.edu/~gbp/mag-ms/mag-ms.PDF

[4] http://listserv.brown.edu/archives/cgi-bin/wa?A2=ind9906d&L=conlang&D=0&P=38635

[5] http://listserv.brown.edu/archives/cgi-bin/wa?A2=ind0410B&L=conlang&P=R11365

[6] http://courses.nus.edu.sg/course/ellmd/Tagalog.pdf

[7] http://www.sultry.arts.usyd.edu.au/LFG98/austro/foley/fintro.htm

[8] http://www.linguistics.ruhr-uni-bochum.de/~himmelmann/LEXCAT.pdf

[9] http://www.linguistics.ruhr-uni-bochum.de/~himmelmann/tagalog_Curzon.pdf

[10]http://www.linguistics.ruhr-uni-bochum.de/~himmelmann/wal_typology_curzon.pdf