"I'm gonna buy me a dog"

Sat Sep 25 08:28:08 UTC 2004

Dear Jo,

I am far from being specialized in English and I'm hence not very
qualified to answer to the 'English component' of your question.
Nevertheless, please allow just a few words, just because the
constructional type you seem to allude to is rather familiar in a number
of other languages, be it Indoeuropean or not. Superficially, we havce
to deal with a 'benebefactive/malefactive reflexive' construction that
is marked by a) the use of a pronominal segment in a positional and/or
formal variant that else occurs for what is often called 'indirect
object', and b) by coreferentiality between the referent of this pronoun
and the 'subject' of the clause. In English, these two parameters are
not immediately visible, as shown by your examples. The positional
left-drift of the pronoun as well as its form both mask the functional
properties of 'me'. Still, iif you re-construct the original (?)
position, you immediately see that the 'role' taken by 'me' is different
from that of an 'indirect object':

I'm gonna bake a cake for me ~ myself.'

If we start from the GIVE-concept as representing a rather prototypical
constructional pattern involving 'indirect objects', it comes clear that
'me' in the clause 'I'm gonna bake me a cake' is NOT an indirect object
strictu sensu. This aspect is even more evindet, if we consider the
first clause ('I'm gonna run me the best race of my life.'). Here, the
for-substitute does not seem to work (at least in my rudimentary
English). This means that the me-constructions would not involve'
indirect objects'. But what IS their nature? AS I said above, the
constructional type is rather common in a number of other e.g.
Indoeuropean lanuanguages, compare Grerman (pseudo-dialectal, region of
Siegen):

Ich rauch'           mir   erstmal  ne   Zigarette.
I      smoke:1sg  me   first         a     cigarette
'I'll first smoke a cigarette (just on my own).'

Trink                 dir            n   Bier
drink:imp:2sg   you:dat   a   beer
'Have a beer!'

This constructional pattern which spreads more and more in colloquial
German perhaps stems from the adoption of parallel strategies from
bilingual Polish-German speakers (immigrants from Poland into the Ruhr
region), as the pattern is famous for many Slavic languages, too,
(though, here the reflexive strategies are slightly different). On the
other hand, we find analogous strategies in many Romance languages. This
would argue in favor of a more general 'cognitive' strategy.

It is important to note that in German, the mir-construction is strongly
related to what one may call 'solipsistic semantics': By this term I
mean the fact that the  mir-construction stresses the 'isolation' of the
primary agent (be it subjective or agentive). The reflexive pronoun
(visible as such in German in the 3rd person only (sich)) hence
functions more like an emphatic element, focusing on its coreferent.
Note that in German the nealry same effect as the mir-construction can
also be yielded by putting stress onto the agent:

(also...) ICH rauch' erstmal ne Zigratte.

This observation goes together with a typology of reflexives that
interpretes them as members of a 'reflexivization scale'. Acccordingly,
the semantics of 'reflexive' constructions starts from highly
S(ubjective)/A(gentive) oriented semantics coupled with corresponding
positional preferences ('near to S/A'). Such patterns yield emphatic
variants of S/A (e.g. I myself). On the other side of the scale, we find
reflexives in Objective (or, if you want: Object) function
(prototypically I kill myself). Again the (now extremely 'distant')
relation between S/A and its coreferential 'O' (in fact usually a
partitive of S/A) is iconically encoded by the corresponding techniques
to distinguish S/A (here: A) from O.

The mir/me-constructions seem to be located somewhere 'in the middle' of
this scale (which by itself is derived from the framework of Cognitive
Typology), hence the term 'middle-voice' seems not too bad. It is what
is called the 'Subjective Version' (SV) in Georgian (or:  sataviso), e.g.

me  v-i-tbob                             sac^'mel-s
I      1sg-SV-warm=up:pres   food-dat
'I warm up the food for me' (~ ? 'I warm me up the food.')

In Georgian, the the pre-stem vowel -i- is a general SV indicator, a
fact that correlates with the well-known distinction between verbal and
lexical reflexivity: In English and in German, we would have to deal
with a kind of lexical SV, whereas in Slavic, a verbal type of SV seems
to prevail. What is interesting in English is the fact that the language
does not use the 'standard reflexive' (myself etc.). In his posting,
Larry has convincingly argued that a phrasing like

> 2b. I'm gonna bake myself a cake.

is even more emphatic than the original phrase (in fact, in Larry's
version, it is the referential entity 'I' which is under emphasis,
whereas in I'm gonna bake me a cake the emphasis also involves aspects
of the verbal action. It would be interesting to see how English would
construe the same phrase with a third person (He's gonna bake him(self)
a cake. ??? - sorry, no competence on my side...).

So, in sum I would dare to adopt the Georgian term (be it sataviso or
Subjective Version) to denote the construction at issue. There surely
are parallels with the standard middle voice, but there also are
differences which should not be put aside by using such a 'cover term'....

By the way: It's rather tempting to interprete the Indoeuropean Middle
Voice endings as a conglomarate of Personal Agreement markers and an SV
marker, e.g. (secondary endings):

            Active      SV
1sg      *-m          *-m-a:
2sg      *-s           *-s-o
3sg      *-t            *-t-o
3pl       *-nt          *-nt-o

The vocalic suffix *-a:/-o (some kind of ablaut?) would than have
functioned just as the Georgian version vowl -i- , or the Slavic pair
-s'/-sja, as in Russian

sobaka  kusa-et-sja
dog        bite:pres-3sg-SV
'The dog bites  (for itself) / *itself

........

I'll stop here, else I would be at risk to start writing an article -
it's just fascinating stuff!

Best wishes,
Wolfgang

Jo Rubba schrieb:

> Hi, all,
>
> People on our grammar-teaching list have been discussing indirect
> objects. Examples from various English dialects came up, such as "I'm
> gonna run me the best race of my life" or "I'm gonna bake me a cake".
> Are these being analyzed as middle-voice (or similar) constructions?
>
--
Prof. Dr. Wolfgang Schulze
Institut für Allgemeine und Typologische Sprachwissenschaft
Department 'Kommunikation und Sprachen' (Dep. II) - F 13/14
Ludwig-Maximilians-Universität München
Geschwister-Scholl-Platz 1
D-80539 München
Tel.: ++49(0)89-2180-2486 (Sekr.) / -5343 (Büro)
Fax: ++49(0)89-2180-5345
Email: W.Schulze at lrz.uni-muenchen.de
Web: http://www.ats.uni-muenchen.de/wschulze