Completely agree

Document Sample
Completely agree Powered By Docstoc
					Eugene Helimski

Comments to the plenary paper by Juha Janhunen
«Reconstructing Pre-Proto-Uralic typology
spanning the millennia of linguistic evolution» *

On the threshold of the new millennium it is useful and tempting to cast a glance back
on many preceding millennia of linguistic evolution. Together with many other congress
participants, I am thankful to Juha Janhunen for the choice of his plenary topic, for
singling out very exactly the core items of Proto-Uralic (PU) and Pre-Proto-Uralic (PPU)
research, and for defining his approaches to these items. Independently on often
fundamental differences in opinions between Janhunen and me, I must admit that his
review is as qualified as it is thematically and chronologically broad.
         By the way, this thematic broadness of issues which Janhunen addresses in his
paper forces me to move in my comments from one issue to another, even at the expense
of being sometimes declarative rather than argumentative. A possible excuse consists in
the fact that I already had an opportunity to explain my views in more detail in some old
and recent articles and books (cf. Õåëèìñêèé 1982, 2000).
         First of all, I permit myself to repeat or slightly reformulate the following
principal standpoints of the speaker which I completely share:
         the Uralic languages form a genetic unity exhibiting all typical characteristics of
divergent evolution accompanied by areal (areal-genetic) connections between
geographically adjacent branches;
         there is no need to radically revise the comparative methods employed in Uralic
studies, and the so-called revolution stirred up by some colleagues was not provoked by
any actual crisis in our discipline;
         the spread of the Uralic languages is best explained by the model of migration
and diffusion from a relatively small homeland;
         there is and will never be a sound alternative to the concept of family tree, though
it must not be seen as a graph from mathematics consisting only of nodes and stretches;
in this connection I would refer to my model of areal-genetic developments for the Uralic
family which combines the advantages of a tree model and of a comb model (Õåëèìñêèé
         the conception of PU as a «primitive» language can today be dismissed as
obsolete together with the glottogonic explanations stemming from Paavo Ravila and his
followers, which still reoccur in works and handbooks on Uralic historical
         the system of numerals in PU was still incipient and was completed only in the
individual branches; this process can be viewed as a cultural innovation - introduction of
decimal counting - which spread over the whole Eurasia approximately 6,000-7,000

  My comments refer to the full text distributed prior to the congress and published in vol. 1 of its
proceedings. In his presentation at the plenary session on August 8, 2000 J. Janhunen mainly treated the
issues of methodology, doing this much more detailed than in the published version. I can express my
full support to his consequently „conventionalist“ and „counter-revolutionary“ approaches.
years ago: there seem to be no proto-languages younger than this dating that did not have
a system of numerals, and no older proto-languages that did (Õåëèìñêèé 2000: 491-3);
         there is little (and even: no) evidence of Indo-Uralic contacts that preceded the
Aryan influences on Finno-Ugric, but the evidence for areal connection between Uralic
and Altaic cannot be disregarded;
         the languages that may be assumed to have incorporated a significant and recent
non-Uralic substrate are, in addition to Hungarian, Saami, Baltic Finnic, Northern
Samoyedic (especially Nenets and Nganasan) and Khanti. I’d add to this list also some
Selkup dialects (esp. Upper Ket’) as well as the now extinct languages of Merya and of
Zavoloch’e Chud’. (Otherwise the issue of substrate is the topic of my own paper at this
         Janhunen’s suggestion to place the Uralic homeland «fairly far in the south, in
the forest steppes of Southern and Southwestern Siberia» is based upon the idea the «the
common direction of linguistic diffusion along the Uralic belt has always been from
south to north». If we accept the last statement as a general rule, there will be no way of
accounting for the geographical location of Finnic and Saami: moving from South or
Southwestern Siberia northwards will never bring anybody to Fennoscandia. If we do not
accept it as a rule, there remains no justification for the southern localisation of the
homeland. Besides, I see no ways and no need to dismiss the well-known palaeolinguistic
evidence for placing the homeland in the zone of coniferous trees and numerous rivers.
         The estimate of the possible age of PU with a surprisingly early date (9,000 -
7,000 BP) also seems to be based on a contradiction in premises. The line of Janhunen’s
reasoning is as follows: there are hardly more than 150 PU etymologies – Proto-Indo-
European (PIE) has much more - PU is therefore much older then PIE. In fact, however,
the crucial figure 150 came into being in 1980 when Janhunen himself undertook a
radical revision - or, I would say, sterilisation - of Uralic etymologies, crossing out
everything that he found suspicious or not fitting his, exact but very incomplete, set of
phonetic correspondences between Finno-Saamic and Samoyedic. But, as far as I know,
neither Janhunen nor anybody else has ever tried to apply a similar Procrustean principle
to the stock of existing IE etymologies. And there are reasons to suspect that after such
application not much - perhaps not much more than 150 perfectly sterile items - will
remain to poor Indo-Europeanists.
         Undoubtedly, in many cases the Uralic etymologies quoted in the UEW,
especially those supplied with a question mark, are absolutely invalid. And nevertheless,
I were ready to show that at least as much as ca. 300 Uralic etymologies must be
reckoned with even without deviating too much from Janhunen’s set of phonetic
correspondences - and this set does not account for some special, statistically rare
structural types.
         I agree that accepting paradigmatic alternations in a proto-language can become
a dangerous weapon for producing heaps of artificial, formally acceptable but cognitively
worthless etymologies, as the sad experience of (some) Indo-European studies and of
Jorma Koivulehto shows. Does this mean, however, that we must have our eyes closed at
such alternations in order «to keep Proto-Uralic clean of such sources of complication»?
(Naturally Janhunen meant here not the PU itself but our reconstruction of it: PU ceased
to exist so long ago that it cannot be influenced by our successes or mistakes.) Here I see
another facet of Janhunen’s methodology of sterilisation; he applies it though he knows
himself better than anybody that a living language - and PU was a living language -
cannot be structurally uniform and transparent.

        The «Pre-Proto-Uralic» of Janhunen is for me nothing else but Proto-Nostratic,
which can be also named «Pre-PIE», or «Pre-Proto-Altaic», or «Pre-Proto-Kartvelian»1.
The difference consists in the fact that Janhunen approaches the corresponding issues by
means of internal reconstruction, while in the Nostratic studies the emphasis is made on
external comparison. Nevertheless, it is no mere chance that several hypotheses on PPU
suggested by Janhunen have exact counterparts in the works of Nostraticists. I can
indicate some of such ideas: maximal paradigm of vowel distinctions in non-first
syllables which underwent reduction in PU (I am, however, not sure that this process
really led to minimal vowel paradigm which Janhunen postulates for PU); lack of
differentiation between nouns and verbs (which, as Janhunen cautiously but correctly
assumes, is rather due to lack of our ability to differentiate between derivatives of the
same stem); origination of suffixes from independent words. I do not quite understand,
why, when analysing the pronominal origin of verbal and possessive suffixes, Janhunen
hesitates to take the next and obvious step and to postulate the word-order VSO or VOS
in PPU - at least in sentences with pronominal subjects2.
        In his previous works, especially in the dissertation «Glottal Stop in Nenets»,
Janhunen attested himself as a matchless master of internal reconstruction. Does this
mean that we must believe him also when from Nenets (for which all solutions in internal
reconstruction are prompted by and can be checked up with external evidence) he moves
into the depths of primary linguistic evolution? I am afraid that it is not so, and that some
statements go beyond the borderline between internal reconstruction and its misuse. (I
must mention that nothing seems to have discredited the method of internal
reconstruction more than its application to Pre-Proto-Indo-European studies).
        One of my main objection concerns the idea of typological cycle. Undoubtedly,
such cycles exist, but the fact that any agglutinative morphology could only have evolved
from an isolating one does not suggest that it did so within the last three, or five, or even
ten thousand years. In particular, the phase of typological cycle attested in most
contemporary Uralic and Altaic languages (agglutination with elements of fusion) may
be very stable and prolonged. Therefore I simply cannot accept Janhunen’s conclusion
that the previous «typological cycle completed immediately before Proto-Uralic» (what is
the source of such dating?) as well as his reflections and conclusions on typological
similarity and chronological incompatibility between «older» PU and «younger» proto-
languages of the Altaic complex. After all, the «young» Russian language is
typologically much closer to the «old» Sanskrit and Latin than to such «youngsters» as
French, English, or Bulgarian.
        No less arbitrary looks the assumption that PPU «lacked a suffixal morphology
altogether»; here, however, the author tried to find a balance between deductive
conclusions and reality.
        I want also to express my strongest doubts concerning the exercises in searching
for fossilised PPU derivatives and compounds in the PU lexicon, as started by Daniel
Abondolo (most certainly not without a postponed influence by Robert Austerlitz) and
supported now by Janhunen. These exercises are alluring and fascinating, but little
productive. I must side completely with the plenary speaker who finishes his discussion
of such PPU etymologies with the following remark: «We still remain waiting for a
substantial confirmation».

  Later phase of PPU can be identified with Uralo-Yukagir; regretfully, Janhunen does not touch the
problems of Yukagir in his paper, and I also omit the corresponding and potentially very relevant
  Cf. numerous languages (French is one of them) where the positions of nominal and of pronominal
actants in relation to verb are different.
         I am optimistic enough to believe that these substantial confirmation - or, which
is at least as much likely, substantial refutations - will come. But I am also pessimistic
enough not to expect this to happen in the next decades or even during the lifetime of the
youngest participants of this congress. We all can be thankful to Janhunen for his
synopsis showing how little one knows and how much one wants to find out about proto-
languages (not only PU) and their prehistory. But, after all, we are not the first and - I
hope - not the last generation of linguists concerned with PU and PPU issues.


Õåëèìñêèé, Å. À. 1982. Древнейшие венгерско-самодийские языковые параллели
       (Лингвистическая и этногенетическая интерпретация). Москва.
--- 2000. Компаративистика, уралистика: Статьи и лекции. Москва.