6.305 Focus systems and Middle Egyptian

[The Linguist List]

----------------------------------------------------------------------
LINGUIST List:  Vol-6-305. Sun 26 Feb 1995. ISSN: 1068-4875. Lines: 348
 
Subject: 6.305 Focus systems and Middle Egyptian
 
Moderators: Anthony Rodrigues Aristar: Texas A&M U. <aristar at tam2000.tamu.edu>
            Helen Dry: Eastern Michigan U. <hdry at emunix.emich.edu>
 
Asst. Editors: Ron Reck <rreck at emunix.emich.edu>
               Ann Dizdar <dizdar at tam2000.tamu.edu>
               Ljuba Veselinova <lveselin at emunix.emich.edu>
 
-------------------------Directory-------------------------------------
 
1)
Date:    Wed, 22 Feb 95 22:25 PST
From: benji wald                           (IBENAWJ at MVS.OAC.UCLA.EDU)
Subject: (COPY)  focus systems and Middle Egyptian
 
-------------------------Messages--------------------------------------
1)
Date:    Wed, 22 Feb 95 22:25 PST
From: benji wald                           (IBENAWJ at MVS.OAC.UCLA.EDU)
Subject: (COPY)  focus systems and Middle Egyptian
 
 I have sent you a copy of my reply to Vincent de Caen because I thought
it might be of more general interest to the list.  Vincent sent a
query to the list, as follows:
QUOTE:
In Middle Egyptian, there is a unique contrast.  There is a particle
iw that most naturally is analyzed as an overt complementizer.  We
obtain the following contrast:
 
sdm.f           "when/if/because/etc he hears"
iw sdm.f        "he hears"
 
sdm.n.f         "when/if/because/etc he heard"
iw sdm.n.f      "he heard"
 
Notice that the markedness relations appear backward: iw apparently
marking the unmarked simple declarative.
Q: are there other systems that employ an overt declarative
contrasting with zero (zero having marked semantic effects)?  And
crucially, what is the etymology of such a particle??   UNQUOTE
 
One of the points in my reply is that the notion of markedness Vincent
alludes to is linguocentric (or whatever the word is).  The way I try
to demonstrate that may be of more general interest, since various forms
of this notion occur on the list from time to time, and are part of our
Greek-Latin hangover -- which we do try to guard against -- when we are
aware of it.
 
The other point that I think is of general interest, and provides a
context for queries I could ask of particular readers has to do with
"scope of focus" systems in languages.  My knowledge of Middle Egyptian
is much less than my knowledge of the other languages I discuss, but
I have read descriptions which suggest (to me) the comparisons I make
below.  So, I would certainly appreciate hearing from anyone who knows
more about Middle Egyptian, and particularly the diachronic behaviour of
'iw.  I would equally appreciate hearing either criticism of the
focus systems of Gikuyu, Yoruba and Mende (and/or related languages)
as I discuss them below, and about other relevant focus systems anywhere
in the world -- and certainly in Afro-Asiatic and Nilo-Saharan languages
in the Afro-Asian area.
-------------------------Text-of-forwarded-mail--------------------------------
 
Date    Wed, 22 Feb 95 21:47 PST
To      decaen at epas.utoronto.ca
>From    benji wald                           (IBENAWJ)
Subject  focus systems and Middle Egyptian
 
Hi, Vincent.  Again, with the Egyptian thing, you ask a good question, in my
opinion -- good, because it exposes the insufficiently relativistic view of
what is currently (perhaps) considered marked and what is unmarked
(unsurprisingly iconic with morphological and syntactic phenomena in
European languages).
 
The relevant dimension in this case might be characterised as "scope of focus",
particularly of constituents of a predicate (VP), and in various ways it
inters ects with various other dimensions in different languages, as you'll
see comparisons I make below.  I'll give some examples in
relevant languages about how these systems work, but then, for
the most part, I'll just give schematics by English glosses to
highlight the main syntactic strategies involved, and not have
to worry about distorting the orthographies used for the
languages of examples.
 
To my knowledge, the prime case is for comparison with Middle Egyptian is
the Bantu area of Interior Kenya, extending from the Kamba/Gikuyu area
to the Southern area of Lake Victoria, e.g., Gusii (East) and
Haya (West).  The curren t southern limit is the Chagga group in
the Kilimanjaro area of Tanzania, which still has morphological
traces of the system I am about to describe.  Gikuyu is
illustrative of the basic system.
 
In this system, when a proposition is asserted rather than
presupposed, the predicate complex is introduced by the bound
morpheme ni~, e.g.,
    ni~-tu~-mw-enda   (FOC-we-her-love)  "we love her"
In Gikuyu, the orthographic convention is that a nap to be written
vowel symbol to show that the vowel is a phonemically distinct mid high vowel
(NOT nasalisation).  At issue is the initial particle ni~-, glossed as FOC.
What it does is indicate that the SCOPE of the FOCUS OF ASSERTION is the
entire clause (i.e., the entire PREDICATE).  Thus, it is not only used in
this role in speech, but is the citation form elicited from
speakers in isolation, a s when the fieldworker asks a speaker
how to say "I love her, you love her, etc. etc."
 
The elicitation necessitates an even focus on the entire predicate.
Many Bantu languages have special auxiliaries to accomplish this purpose,
instead of the preposed FOC marker.  They are similarly elicited by the same
fieldwork techniques, giving the naive analyst the impression that they will
turn out to be more frequent in spoken discourse (by analogy to the
European --  or Swahili -- stimulus sentence) than they turn out to be.
 
The point about frequency is directed against your assumption that
Middle Egyptian 'iw sdm-f etc is characterisable as "the unmarked
simple declarative" -- probably because either you assume that this is
a universal category or because you are using some koine linguistic
terminology to question that notion.  Once you reflect that it is essentially
a taxonomic label derived from a tradition used to analyse certain European
languages, you may become more cautious in applying it to other languages.
Also, the addition of the word "unmarked" in the label is a recent
importation, and has no theoretical (or even comprehensible) status
othe r than to indicate that the construction is morphologically
simpler than some other relatable cons tructions.  That makes it
a formal descriptive term, and absurd for the Middle Egyptian data.
 
As one would expect from the above characterisation of the function
of ni~-, it is not used in wh questions, since, then, the focus
is not spread evenly across the whole proposition but is
concentrated on the question word,  nor in answers to wh- questions, where the
focus is on the information requested by the wh- word, e.g.,
    u-uma ki~ (you-come.from where)  "where are you coming from?"
     ny-uma ru~u~-i (I-come.from river)  "I'm coming from the river"
As in the above examples, the wh-word or its answer comes last in the
proposition, because this is the position of maximum focus (ceteris paribus).
 
As is the case with the languages which use auxiliaries to even the focus of
the entire proposition, ni~- is used as a strategy to COUNTERACT the
attraction of maximum focus to the last word.  (Below I will
explain the special case of a wh-word as subject).
 
Similarly, it is not used with negation or in subordinate clauses,
where the FOC marker would contradict its function to spread
focus across the entire proposition (necessarily including the
"main" clause), or in consecutive ("and-then") clauses (which
are distinctively marked, and cannot have focus ind ependent of
the initial clause).  With respect to consecutive clauses, it is
worth pointing out that while focus is often used to highlight "new"
information in propositions as in the wh-question/answer examples, "new"
in the context of the initial clause of a temporal sequence of
clauses means a new section of discourse, episode, or whatever
you want to call it.
 
The origin of ni~- in these languages is undoubtedly the same as
the Proto- Bantu "equative copula" *ne (of Pre-Bantu origin,
found elsewhere in Niger- Congo), still used in the same
function in most Bantu languages. This particular equative
copula is, I think, originally impersonal, i.e., had n o
subject originally (so that it would translate into English
 as "it is"), but ca n be analysed as allowing a subject
(actually the "topic" of the predication) in all Bantu languages, cf.
 
Gikuyu:  mu~gu~nda ni~ mu~nene (field FOC big)  "the field is big."
 
Thus, ni~, called an "equative copula" by linguistic tradition, is a
marker whi ch focuses on the assertion of the predicate as a whole.
 
Considering now the "special case" of a wh-subject, as in "WHO
broke the plate?", while some Bantu languages (but relatively
few) put WHO in word final position, Gikuyu, like many Bantu
languages, uses a "copula" construction , e.g.,
 
 n-u~ u~-roragire thani (FOC-who s/he-broke plate)  WHO broke the plate?
 
The literal translation is: "it.is-WHO (that) broke the plate?" But
the strategy is more precisely to allow the "subject" to be in
focus by recasting it as a nominal predicate marked by the focus
marker ni~.  The associated "embedded" clause, of course, cannot
have a FOC marker, just as a relative or other dependent clause
cannot.  Also, it is obvious that the "dependent" clause
contains presupposed/old, NOT asserted/"new", information (even though
the guilty party will undoubtedly say "WHAT plate?").  It is a
common Bantu strategy, even among languages which do not (or no
longer?) use ni~- as a verbal predicate focus marker.
 
The generalisation of this function from nominal predicates to verbal
predicate s seems to be an areal innovation, and cannot be
reconstructed for Proto-Bantu. However, the logic of the system,
as indicated by my comment above comparing the focus function of
ni~ to special auxiliaries in other Bantu languages, is probably
of Pre-Bantu origin (and is wider in  Niger-Congo), making this a
more natural form of grammaticalisation than it would be in some language
families, including you-know-what.  That is, if we try to translate
the grammat icalisation literally we get the misleading "it is
(the case that) I love her".  But such an English locution,
while possibly (counter-)assertive, does not seem appropriate if
the proposition "I love her" is NOT presupposed. Thus, in
English (among many languages) it is difficult to see how
such a construction could become "unmarked" to mean something as
simple as "I love her".  However, in languages where focuses of
assertion tend to be  mor phologically marked, at least to
control their scopes, it is easy to see the ge neralisation from
nominal predicate assertions to verbal predicate assertions.
 
Yoruba, among many coastal West African Niger-Congo languages, has a
comparable FOC marker *ni* (morphophonemically denasalised to l- in
predictable phonological contexts), probably cognate with Bantu *ne.
In contrast to Bantu *ne it is postposed to the word or phrase
that it focusses on.  Also in contrast to Bantu, it combines
with preposing of the focussed constituent in all cases.  Thus,
WH questions ar e similar to Indo-European because of the word
order strategy, e.g.,
   who FOC s/he see him   (who did s/he see?)      NB subject-marking oblig
   who(m) FOC s/he see    (who/m did s/he see?)   NB object "MOVED"
   he FOC I see      (he's the one I see)   NB "he" in emphatic (polysyllabic)
                                                           form
   Nom-see FOC I see him  (I did see him/I merely saw him)
And, of course, the uses that look like "copulas"
   him/her FOC  (it's him/her//s/he's the one)  NB "him/her" in emphatic form
   he broke it FOC (it's the case that he broke it -- that's what he did)
 
It is evident that preposing and morphological FOC marking is stronger (i.e.,
a more favoured strategy), and FOC on the final word/phrase in a
proposition is much weaker (if operant at all), than in Bantu.
 hus, the last example is parallel in structure to the Gikuyu
innovation, but it does not have the same contexts of use.  That
follows from the absence of automatic focus on the last word of
a proposition in Yoruba (and other West Niger-Congo languages),
so that the FOC marker is not simply used to even out the scope of
focus to the entire proposition, but to RAISE its focus for
emphatic purposes.  Finally, note in view of what I suggested
about the "original" use of Bantu *ne, that Yoruba ni is
"impersonal".  As in Bantu, such a "copula" (if one insists on
  calling it that) is distinct from other "copulas" which may
have (and/or be inflected for) nominal subjects, as if they were verbs.
 
Maybe more interesting to compare with the Gikuyu use of the focus marker
is the use of the focus marker in various Mande languages, e.g., Mende.  To
begin with, Mande languages have SOVX order, in contrast to the SVO order
of Yoruba etc., and the more flexible but still marked tendency toward SVO
in Bantu.  Mende is illustrative of the Mande focus marker.  As in Yoruba,
the FOC marker is postposed to the constituent focussed upon.
 
 he plate break-FOC (he broke the plate)
  he plate FOC break (he broke THE PLATE)
  he it break yesterday FOC (he broke it YESTERDAY)
  he what FOC break (what did he break?)
  child FOC (this is the/a child)  NB there is a deictic element in
                                                  this context
 
The most interesting example for comparison with Gikuyu ni~ is the first.
The Mende FOC does not focus on the verb, but is the neutral
"normal" proposition, with even focus on the entire predicate,
just like with prefixed Gikuyu ni~. Unfortunately, I don't know
enough about the history of Mende or Mande to understand how the
construction originated, but I don't doubt that the interaction
of morphological FOC and the use of word order was involved,
as in Gikuyu. Probably on the right track is the notion that in a
"neutral" sentence the focus is "naturally" on the verb, as the
head of the entire predic ate. Also interesting is that, as in
Gikuyu, but unlike Yoruba, the FOC marker does not disturb word
order when a constituent is focussed on, cf. the Gikuyu
examples given above,
   FOC-he(/who) he-broke plate  (it was he that broke the plate/
                                 who broke...?
   he FOC-he-broke plate     (he broke the plate)
Mende, having the verb in final position (with respect to a nominal object),
"innovates" (from the perspective of Gikuyu) by allowing the FOC marker to
become mobile.  There may be implications in this idea for a shift from SVO
to SOV in Mande -- but let's not get into that.  In the final analysis, the
 Men de FOC "postposition" is a morphological counterpart to
English suprasegmental stress.   It can mark any constituent
without affecting word order.   The counterpart to the
obligatoriness of the Mende FOC marker (in relevant
constructions) is that a primary stress must fall somewhere in the
English clause.  If there is no "unusual" focus in Mende FOC
marks the verb.  The English equivalent is that the primary
stress falls on the verb -- and here the implication is that
"unusual" in English would be a stressed object or some
non-verbal constituent in the verb phrase as opposed to an
unstressed object -- usually an anaphoric pronoun (given a
transitive sentence)
 
If I remember correctly, 'iw is one of the Middle Egyptian
"copulas", not more blandly a "particle" -- or
perhaps the latter is a categorisation of its function in the
relevant context .  Therefore, a rationale and perhaps even an origin
similar to Gikuyu ni~- is likely.  That is, we are dealing with an areal
phenomenon of considerable antiquity in a concern for indicating the
scope of focus in predications.
FLASH;  I just checked and I do remember correctly, here are some of
the use of 'iw:
 'iw-n m pr-f (FOC-we Loc house-3s)  we're in his house
 'iw-f n-i (FOC-3s to-1s) it's mine
 'iw-f ?nx-w (FOC-3s life-Adj) he's living/alive
 'iw-f r s'dm (FOC-3s up.to hear) he's gonna hear
 'ir s'sp-n-'i ... 'iw di-n-'i xt Hmw
  if  hurry-Asp-I ... FOC make-Asp-I retreat cowards
(if I had hurried ... I would have made the cowards retreat)
   Notice which clause of the conditional pair is marked by 'iw.
 
Also, Middle Egyptian does not "move" WH words, but leaves them
"in place", which seems to be final if they are not the subject,
e.g.,   iry-i m  (do-1s what) what should I do?
This makes the Middle E focus system (if that's what it is) more
similar to the Gikuyu/Bantu system than to the West African ones
(represented by Yoruba above).  It differs in that 'iw has locative
uses, as in the first ME example (and maybe the "gonna" one), while
the Niger-Congo focus markers do not have locative uses.
 
An interesting point in this context is the Niger-Congo examples
I have discussed are tone-languages.  Larry Hyman, who has done
some of the important and pioneering studies on African focus
systems, has suggested that their morphological focus systems
can in some way be understood as ways of avoiding overburdening
their prosodic systems. Whether that suggestion turns out to be
sufficient or not, if Middle Egyptian turns out to have a
similar system in motivation or origin, questions might be
raised about the function of tone in Early and/or Middle
Egyptian.  That would certainly be quite interesting, and novel,
as far as I know.  (Currently, Afro-Asiatic languages below the
Sahara are tonal, those above are not. I know nothing but would
be interested in knowing the details of arguments about how that
happened.)
 
I also think I remember something in Gardiner's grammar of
Middle Egyptian about the loss of this construction in Late
Egyptian (and, of course, Coptic).
 
It is thus reminiscent of the loss of the system in Chagga (where
morphological traces indicate that it formerly was used in a way
similar to languages to its north).  In fact, I think Gardiner
mentioned something about 'iw becoming a marker of certain types
of subordinate clauses in Late Egyptian, so it would be
interesting to check if that is so, and if so, how and why such a form of
evolution occurred.
 
I am going to try to post this to the ling.list as well in the
hopes that my discussion will elicit comments/examples from
other languages by other readers, particularly with respect to
focus systems in other languages in the Afrasian area, or
specifically with respect to use and evolution of the uses of
'iw and the 'iw sdm-f construction in various stages of Egyptian.
 -- Benji
 
--------------------------------------------------------------------------
LINGUIST List: Vol-6-305.