| AUTHORS: Mallory, James; Adams, Douglas Q.
TITLE: The Oxford Introduction to Proto-Indo-European and the
PUBLISHER: Oxford University Press
David Stifter, Institut für Sprachwissenschaft, Universität Wien
In 1997 James Mallory and Douglas Q. Adams edited an impressive _Encyclopedia of
Indo-European Culture_ (EIEC). The book offered a treasure trove of factual
information on the prehistory and material culture of peoples speaking
Indo-European languages, in a large and heavy volume that combined lexical data
and archaeological facts in a framework of cultural-historical interpretation.
Despite some criticism EIEC was a major advance by the mere fact of having been
the first such undertaking in Indo-European studies for almost three quarters of
a century. Almost a decade later the two scholars have produced a handbook
version of it. The new book may be characterized briefly as EIEC without the
archaeology and without the pictures, focusing almost entirely on the
reconstructable lexicon and on the evidence it provides about the people, their
physical, social and spiritual lives, and about the environment in which they
lived in the Proto-Indo-European (PIE) period 6–8 millennia ago. The target of
the book are ''general readers, students [...] as well as professionals in
disciplines such as archaeology who need to deal with the early Indo-Europeans'',
and ''linguists interested in refining, challenging, or adding to our
understanding of Proto-Indo-European'' (p. xxiii).
The book consists of 26 chapters. The first four chapters concisely introduce
the Proto-Indo-European language and the Indo-European (IE) languages, the
principles of comparative Indo-European linguistics, the methods of
reconstruction, and the grammar of the reconstructed proto-language. This
amounts not to a full grammar of PIE, but to a very broadly outlined overview of
the reconstructable grammatical categories. Chapter 5 is concerned with internal
and external relationships of Indo-European. Among the latter, the possible
genetic relationship with the Uralic language family is mentioned. Chapters 6–7,
perhaps the most insightful chapters of the book, are devoted to the methodology
of linguistic palaeontology, the proper subject-matter of the book, and they
conclude the introductory first part of the book. Ch. 6 ''A Place in Time''
provides a lucid exposition of the methodological problems encountered in trying
to date PIE and to determine the approximate homeland. Under close examination,
neither linguistics nor archaeology offer reliable methods that would allow an
unambiguous answer to these questions. Ch. 7 ''Reconstructing the
Proto-Indo-European'' is not – as the title might suggest – dedicated to the
physical appearance of the speakers of PIE. Rather, the authors elaborate on the
conditions that have to be fulfilled in order to ascribe a specific
reconstructed lexical item to PIE or to an intermediate stage sometime after the
breakup of PIE but before the emergence of the individual IE branches and
languages. A particular strength of the book lies in the fact that the authors
usually tell explicitly if a reconstructed word belongs to PIE or to a younger,
intermediate subnode among IE languages (pp. 109–111). What model of subgrouping
one chooses for IE languages has far-reaching consequences on how one has to
assess the evidential value of a specific feature attested only in a subgroup of
languages. The presence of a word or a semantic field in a narrowly delimited
region can be an indicator for a substratum there (e.g. the distribution of
farming vocabulary in IE languages).
Chapter 7 sets the stage for the second part of the book and highlights the
practical limitations that will be encountered again and again in the ensuing
chapters 8–26. They, the heart of the book, give accounts of the reconstructable
PIE lexicon, arranged by semantic fields, covering topics like fauna, flora,
family and kinship, clothing and textiles, food and drink, space and time,
emotions, mythology, religion, and many more. The arrangement follows C.D.
Buck's (1949) _Dictionary of Selected Synonyms in the Principal IE Languages_.
Each chapter is dedicated to one semantic macro-field (e.g., ch. 13 ''Hearth and
Home''), which is further broken down into sub-fields (e.g., 13.1. ''Dwelling'',
13.2. ''Construction'', 13.3. ''Proto-Indo-European Settlement''). For each field
all pertinent reconstructable terms have been excerpted from EIEC. The
sub-chapters are accompanied by boxes that contain the reconstructed terms
mentioned in the text, plus a translation of these (not always grammatically
fully explicit), plus a handful of attestations. While in the boxes the attested
forms are taken only from Modern English, Latin, Sanskrit or Greek, in the main
text the reconstructions are supported by one cognate form from each branch of
IE. This is a much larger range of languages, still not exhaustive, but
illustrative for an English-speaking audience who are evidently the prime
addressees of the book. English lexemes are regularly employed as Germanic
representatives of IE reconstructions, and it is not infrequent that Germanic
examples are only taken from Modern English; an unusual procedure for a handbook
in historical linguistics. But the present book is not to be mistaken for an
etymological dictionary of Indo-European.
Summarizing sections at the end of each chapter give a well-balanced and sober
assessment of the foregoing material. In these recapitulating sections,
comparison of the PIE situation is regularly made to Proto-Uralic. This gives
interesting contrasts and agreements. Every chapter is closed by a succinct list
of fundamental secondary literature. The last 250 pages of the book are taken up
by various appendixes (including PIE–English and English–PIE word lists), a
reference section, and subject and word indexes.
Throughout the book, one can find noteworthy information concerning the realia
of IE archaeology in the recapitulating sections. I want to highlight two
observations: Reconstructable terms referring to the fauna and flora have played
an important role in attempts to determine the original homeland ('Urheimat') of
the Proto-Indo-Europeans. The fact that the names for just over seventy animals
(Ch. 9) and of approximately sixty plants (ch. 10) can be reconstructed with
some certainty for the proto-language reveals how limited our knowledge about
the lexicon of PIE really is and how much must have been lost over the
millennia. Modern research has revealed that on average agricultural societies
have 400–500 names for animals and plants each. The reconstructable IE faunal
and floral lexicon only amounts to ca. 15% of this. The amount of loss that must
have occurred is enormous, and there is no reason to think that the loss in
other semantic fields was considerably smaller.
Lexical items connected with 'wheeled transport' have ''often been regarded as
one of the most diagnostic semantic fields in the reconstructed lexicon'' (p.
252), because from the well-established vocabulary for 'wagons' and 'wheels' in
the proto-language it can be inferred that PIE as a yet undivided linguistic
entity must have participated in the technical revolution that transformed
transport and locomotion in the middle of 4th millenium B.C. All attempts, like
– most notably – Colin Renfrew's, to assign the IE proto-language and the
emergence of the individual branches a much earlier date than that are therefore
by implication falsified. If transport terminology had spread across the IE
world after the breakup of the proto-language, this would be recognizable by
deviant sound correspondences, the unmistakable diagnostic tool of loan
relationships as opposed to genetic inheritance. A well-supported terminus post
quem for the break-up of PIE can thus be reached. In the context of 14,
''Clothing and Textiles'' the authors draw attention to another possible criterion
for dating Proto-Indo-European: the presence of the words for 'sheep' and 'wool'
give evidence for IE acquaintance with woolly sheep. It appears on
archaeological grounds that the economic exploitation of sheep with sufficient
wool were not known before the 4th millenium B.C. (p. 238).
Authors of lexicographical works – and this book is basically a lexicon – easily
run the risk of producing endless lists of lexical items whose propensity to
conveying boredom is directly proportional to their length, but this is not so
in this case. Although chapters 8–26 ultimately amount to a very long
presentation of lexical material, all chapters are well written, and indeed the
pleasingly readable style of the whole book must be applauded. Having said this,
the rest of the evaluation will be much less favorable.
Various types of errors catch the eye of the informed reader, both factual
mistakes and seemingly small errors and typos that are the result of a
noticeable carelessness and sloppiness in the production of the book. I only
want to cite a few representatives of these:
On pp. 7 and 37 a fragmentarily attested language in the west of the Iberian
peninsula is called ''Lusatian'' instead of Lusitanian (correct on pp. 13–14).
Lusatian is rather a variant name for the Slavic language Sorbian spoken in
Germany that, incidentally, is missing in the list of West-Slavic languages on
p. 26. And on map 1.2 ''Surviving IE groups'' (p. 9), Armenian has been forgotten.
In a book that is also aimed at a non-specialist readership it would have been
appropriate to point out that the authors' decision to reconstruct four
laryngeals for the proto-language instead of three is a marginal position in the
discipline (p. 56). Albanian h in the onset of words like herdhë 'testicle',
allegedly < ''*h4org'hiyeh2'' (p. 55), is no good indicator for the presumed
existence of a fourth laryngeal *h4. The attestation of Albanian sets in only as
late as the 15th century, and its comparative analysis still lags behind most
other branches of IE. This renders Albanian a very weak basis for far-reaching
theories about PIE. In some cases the Albanian sound h may rather be due to
secondary spread, as evidenced clearly by hark 'bow', which is a loanword from
Latin arcus 'id.', while in other cases there are different and more convincing
PIE sources for h available. For example, Albanian ha 'to eat', allegedly
continuing a PIE root h4ew (p. 255), could better be compared with Old Indic
khad- 'to chew' and be traced back to PIE kh2ed 'to chew' (Matzinger 2006: 80).
Similarly idiosyncratic is the treatment of the formal categories and semantic
dimensions of Indo-European. To name only the most striking points in table 4.3.
(p. 57), which purports to illustrate the ''Basic case endings of the
Indo-European noun'': conspicuously absent here is the ending of the genitive
singular in *-es; the original shape of the accusative plural was *-ms, not
*-ns; a dative plural ending * mus is idiosyncratic in the extreme (cp., for
example, Tichy 2000: 65–66, Meier-Brügger 2000: 183–185 for a more mainstream
account of the matter). And in the verbal domain: on p. 63, doubt is cast on the
PIE status of the categories subjunctive and imperfect (note also that the
Celtic languages are not conceded an imperfect by the authors), while at the
same time the possibility of a separate category future is considered for the
proto-language (cp. again Tichy 2000: 80 ff., Meier-Brügger 2000: 153 ff.).
It is particularly confusing, especially for non-specialist readers, that in the
introductory chapters typos disrupt reconstructed PIE forms, doubtlessly due to
incompatibilities in the use of special fonts in the production of the book.
Table 4.1 presents the PIE phonological system, but those sounds, for which no
letter exists in the standard English alphabet, are misrepresented: 3, C, i and
an empty slot stand for syllabic r, l, m, n respectively (p. 55). On any page
picked at random it is possible to discover minor and major errors that could
have been easily avoided: for Old Irish aithe 'foster-father' (p. 211) read
aite; Old Irish oa, allegedly meaning 'young' (p. 205), is actually the
comparative of oac 'young'; Old Irish bolgr 'sack' (p. 230) was apparently
furnished with an Old Norse case ending, correct is bolg; and the Old Irish
preposition for 'in' is i, followed by nasalization, not in (p. 290). It is most
annoying that these ghost-forms have also found entrance into the word index. A
double error involving false friends occurred on p. 271: Old Irish nem does not
mean 'gift', but rather 'heaven'. What the authors probably had in mind as a
derivative of the PIE root nem 'to give, take' was neim, which indeed means
'Gift', although not in English, but in German (i.e. English 'poison').
When one encounters hundreds, if not thousands, of those small signs of
carelessness, it comes as no surprise if one starts to wonder whether the same
amount of unreliability underlies the book as a whole. And indeed this is the
impression that one gets regarding its etymological sections. Necessarily, given
the scope and the aim of the book, the authors could not explain how they
arrived at the reconstruction of the individual lexemes. This will leave one
looking in vain for detailed etymological analyses. Phonological or
morphological problems receive only the most superficial attention. For example,
it is vaguely stated that Latin vehiculum and Sanskrit vahítram, both 'wagon',
differ from e-grade Old Irish fén 'wagon', Tocharian B yakne 'way, manner' and
from o-grade Old English wægn by a ''different suffix''; Old Church Slavonic vozŭ
and Greek ókhos are simply said to be ''still another formation'' (p. 247).
Alleged word correspondences tend to have the value of mere root equations:
readers who are not trained in IE linguistics (remember that they are the prime
addressees of the book) are thus led to believe that Russian vojë 'pole, shaft',
Hittite hissa- 'pole, shaft, thill', Avestan aēša- 'plough, pair of shafts',
Sanskrit īshā- 'pole, shaft', Modern English oar and Greek oiḗïon 'tiller, helm,
rudderpost' all regularly continue PIE *h2/3éih1os 'shaft of a wagon' (p. 249)
and not different derivatives. (Note that English oar, for which Proto-Germanic
*airō can be set up on the basis of Finnish airo, can hardly belong here, unless
a very early date is assumed for rhotacism.) Another example: ''*k'(o)nid- 'nit,
louse, egg''' is said to be ''well attested with cognates in Celtic (e.g. OIr sned
'nit'), Germanic (NE nit), Baltic (e.g. Lith glìnda), Slavic (Rus gnída), Alb
thërije, Grk konís, and Arm anic'' (pp. 150–151). I wonder if a presentation of
the facts like this, where no regular sound correspondences between the words
can be discerned, really helps to strengthen the confidence of laypersons in IE
studies as a rigorous science, and does not rather add fuel to an impression of
arbitrariness in the discipline. The editors could at least have remarked that
in taboo words like 'nit, louse' irregular phonetic distortions must be expected.
Sometimes the reconstructions are simply wrong. For Albanian rreth, pl. rrathë
'hoop, tire' and Tocharian B retke, Tocharian A ratäk 'army', the authors set up
a pre-form *róth2ikos, a form that is utterly unable to explain either of the
words. First of all, the authors have fallen victim to their own sloppiness: in
EIEC they had still reconstructed *rothik'o- for Albanian, where k' accounted in
a regular manner for Albanian th. In the present book, the tectal sound has – no
doubt by oversight – lost its diacritic, thereby turning into a plain velar,
which, however, would have resulted in Albanian k. This is not to suggest that
the reconstruction proposed in EIEC was correct – far from it. The alleged
development *roth2ik'o- > Proto-Albanian *rratitha- > *rretithë > *rret'thë >
rreth (uel sim.) requires a spontaneous syncope of the second syllable for which
there is no evidence elsewhere in the language; and the preform is unable to
explain the irregular plural rrathë, which has been ignored by the authors. For
an explanation that tries to account for all forms (i.e. *roth2ih2-) see fn. 3
in Stifter (forthcoming). The case of the Tocharian words allegedly continuing
*róth2ikos is likewise problematic. If the established soundlaws of Tocharian
are applied, the preform should have resulted in Tocharian B **recake, Tocharian
A **racäk < Proto-Tocharian **rëcäkë.
The gaps in the reference section are eloquent: scholarship in the German
language has been and still is very prominent in Indo-European studies, not
least in Indo-European archaeology. Therefore it comes as a surprise that
important German contributions like Zimmer 1990 and Schmitt 2000, which touch
upon the very heart of this book's subject-matter, are missing from the
bibliography. Important reviews of EIEC like Polomé 1998 and, particularly
substantial, Zimmer 1999 (a review of 58 pages) are likewise absent.
Furthermore, several names have been mis-spelt: ''Gvozdanivic'' for Gvozdanović,
''Kurlowicz'' for Kurylowicz [editor's note: the l in both versions should have a
bar through it], ''Mazjulis'' for Maziulis [editor's note: the correct version
should have a hacek over the z], ''McKone'' for McCone, ''Meier-Brugge'' for
Meier-Brügger, ''Sturteyant'' for Sturtevant. The authors have not only introduced
ghost-words, but also at least one ghost-book: ''Katz, J. (2003). Studies in
Indo-European Personal Pronouns. Göttingen: Vandenhoeck and Ruprecht'' (p. 576).
The only available version of Joshua Katz's study remains the mircofilm edition
of his PhD-thesis 1998.
Lest I appear as a relentless critic, I also want to make two constructive
suggestions. The first one concerns the list of poorly attested IE languages
(pp. 13, 36–37). In this context reference could also be made to attempts at
identifying languages that have not survived in any directly accessible manner
(i.e. either as spoken languages or transmitted in a corpus of written texts),
but whose existence has been inferred indirectly either on the basis of possible
loanword layers in other branches, like the language tentatively called
'Temematic' (because of the postulated sound-changes concerning the obstruent
series PIE TEnues > MEdiae and PIE Mediae Aspiratae > Tenues) by Holzer 1989, or
through the interpretation of toponomastic evidence, like, for instance,
possible IE languages in the Alps ('Ostalpenblock' or 'Eastern Alpine IE') and
in Central Europe (Pannonian) (see esp. Anreiter 1997 and 2001). Irrespective of
what one thinks of these postulated languages (many of these attempts have met
with criticism), the picture of the extent of IE remains incomplete without
these marginal entities, as long as the methodology by which they have been
postulated cannot be shown to be obviously flawed.
In chapter 20 about ''Indo-European Flora'', *ghabhlo/eh2 'fork, branch of a tree'
is introduced as a word exclusive to Celtic and Germanic, the most westerly
branches of IE (p. 160). But this word is quite transparently a derivative with
the instrumental suffix -lo/eh2- from the root *gheHbh 'to take'. The meaning
'fork', from which 'branch' must derive secondarily, demands a literal
translation 'instrument for taking', which presupposes the verbal meaning 'take'
for the root. This meaning is only to be found in Celtic (pace De Bernardo
Stempel 2005). In consequence, Germanic *gabla/o- must be a loan from Celtic,
and the word does not belong in a floral context, but has been transferred there
So, what impression remains of the book, after one has read it? The ultimate
introduction to IE linguistics has not yet been written. This book cannot fill
this gap, either, and given the steadily increasing refinement in IE studies
such a book may never be written at all. As a general introduction to the
current knowledge of PIE culture and archaeology this book is of some, but not
of undiminished, value. The amount of concise information about realia and
material culture could have rendered the book a precious supplement to
Indo-European etymological dictonaries like J. Pokorny's _Indogermanisches
Etymologisches Wörterbuch_, H. Rix's _Lexikon der Indogermanischen Verben_, and
other linguistic lexicons that are in preparation at the moment. It is a pity to
say that a good opportunity has been wasted.
A lot of what is said in this book is not in line with standard IE linguistics.
Either the standard methodology is not correctly applied, or standard tenets are
explicitly ignored or contradicted. I am not to be misunderstood: there is
nothing illegitimate in doing so. But deviating from the mainstream and
following a maverick path does not automatically imply that one possesses a
better, because rarer, truth. Untrodden paths may simply lead astray, and there
may be a good reason why most people decide to swim in the mainstream. In any
case, if one does not believe in what amounts to be the majority doctrine in a
discipline, this fact should be made explicitly clear in a handbook geared at a
general audience, and ideally the minority position ought to be argued through.
The most illuminating sections of the book are those that are devoted to
methodology and theoretical questions concerned with the difficulties
encountered in establishing the material culture and the original homeland of
the PIEs. The rest of the book, which by virtue of its subject matter could have
had all the sympathy of this reviewer, is entirely eclipsed by the carelessness
with which the authors went to their work.
Anreiter, Peter. 2001. _Die vorrömischen Namen Pannoniens_. Budapest: Archaeolingua.
Anreiter, Peter. 1997. _Breonen, Genaunen und Fokunaten. Vorrömisches Namengut
in den Tiroler Alpen_. Budapest: Archaeolingua.
Buck, Carl Darling. 1949. _A Dictionary of Selected Synonyms in the Principal
Indo-European Languages_. Chicago University Press.
de Bernardo Stempel, Patrizia. 2005. ''Indogermanisch und keltisch 'geben':
kontinentalkelt. Gabiae, gabi/gabas, keltib. gabizeti, altir. ro-(n)-gab und
Zugehöriges''. _HS_ 118, 185–200.
Mallory, James P. & Douglas Q. Adams. 1997. _Encyclopedia of Indo-European
Culture_. London – Chicago. (=EIEC.)
Holzer, Georg. 1989. _Entlehnungen aus einer bisher unbekannten indogermanischen
Sprache im Urslavischen und Urbaltischen_. Wien: Verlag der Österreichischen
Akademie der Wissenschaften.
Matzinger, Joachim. _Der Altalbanische Text [E]Mbsuame e Krështerë (Dottrina
Cristiana) des Lekë Matrënga von 1592. Eine Einführung in die albanische
Sprachwissenschaft_ [= _Jenaer Indogermanistische Textbearbeitung_ 3],
Meier-Brügger, Michael. 2000. Indogermanische Sprachwissenschaft. 7., völlig
neubearbeitete Auflage der früheren Darstellung von Hans Krahe. Unter Mitarbeit
von Matthias Fritz und Manfred Mayrhofer, Berlin – New York: Walter de Gruyter.
Polomé, Edgar C. 1998''Review article on the IE encyclopedia''. _JIES_ 26, 279–288.
Schmitt, Rüdiger. 2000. ''Indogermanische Altertumskunde. I. Sprachliches''. In:
_Realenzyklopädie der Germanischen Altertumskunde_. Vol. 15, Berlin – New York,
Stifter, David. Forthcoming (prob. 2008).''OPr. kelleweſze 'Driver of a Cart'''.
Tichy, Eva. 2000. _Indogermanistisches Grundwissen für Studierende
sprachwissenschaftlicher Disziplinen_. Bremen: Hempen.
Zimmer, Stefan. 1990. _Ursprache, Urvolk und Indogermanisierung. Zur Methode der
Indogermanischen Altertumskunde_. Innsbruck.
Zimmer, Stefan. 1999. ''Comments on a great book: The Encyclopedia of
Indo-European Culture, Mallory and Adams 1997'' _JIES_ 27, 105–163.
ABOUT THE REVIEWER
David Stifter is researcher at the Dept. for Indo-European Linguistics at the
University of Vienna. His focus of research lies in Celtic languages.