This translation of the German edition first published in 1970, introduces the standard text on the comparative-historical method to an English-speaking audience. After surveying the general principles of diachronic-comparative linguistics, the book uses these principles to analyze the phonological and morphological structure of the Indo-European language group. Each section of the book has a detailed bibliography, so readers can progress from the general overview to a more in-depth examination of particular topics.
Szemerényi is, of course, famous for his eponymous law, which, though it seems obvious to the point of triviality today, apparently cleared up a considerable amount of genuine confusion at the time. Unfortunately, it seems like that was the only positive thing he ever accomplished in his career—this book, which variously ranges from wildly anachronistic to apparently deliberately unhelpful, certainly isn't another.
The biggest issue by far (but not the only one) is that Szemerényi outright refuses to accept laryngeal theory—a position that was already weird and untenable in 1970, when the first edition was written, and became fully embarrassing long before this edition was produced in 1994. By then he had obviously spent most of his career on the defensive, too obstinate to change his mind in the face of universal disagreement anymore, and it shows in his writing: he can't get away with leaving out laryngeals entirely in what, after all, pretends to be a usable introduction to PIE, so he very briefly covers them (poorly, dickishly), pretending they're an arbitrary and inadequate patch on a non-problem, and then spends dozens of pages pointing out alleged issues with them in the pissiest way possible—this section is so unconvincing, in fact, that it's a testament to his pig-headedness he didn't manage to change his own mind then and there, and particularly anything to do with Hittite is so weak it can only be seen as mean-spirited. He never uses laryngeals himself after this, anyway, and unfortunately, they're just too fundamental to the modern understanding of PIE for a reconstruction that relies on an ill-defined schwa indogermanicum to have much value to the modern student, and there consequently isn't a single part of the book that's really salvageable: the section on phonology is necessarily a mess (not helped by Szemerényi's additional refusal to recognise the palatovelars: he holds, wrongly, that they're an innovation of the satem languages), that on morphophonology makes an inevitable hash out of things (disyllabic roots!), those on morphology are at best fully a century out of date and in general a lot more concerned with grinding very specific axes than with communicating fundamentals (for the verbs, in particular, his rejection of laryngeals is a great slur against the conservatism of Greek's εἰμί paradigm).
If there were nothing else wrong with this book, that would be enough of a reason to relegate it to the "historical curiosities" bin along with Szemerényi himself; it is not the case that there is nothing else wrong with the book. This issue is so fundamental, though, that it seems pointless to mention other very basic factual problems (like the off-hand claim, no examples given, that full-grade *kʷe alternates with zero-grade *ku (!!!)) or his insistence on terminology that was outdated long before the first edition got published ("Aryan" for Indo-Iranian is common enough among racists, but I'm still not sure what "Old Indic" is supposed to be) or his refusal to translate quoted materials in French, German, or Latin (!) that seems at odds with any actual desire to inform.
Surprising and genuinely upsetting. Szemerényi wasn't just a poor scholar, he was a bad teacher and just comes off as an unpleasant person as well.
Eines der ganz wenigen Bücher, die ich von ehemaligen Dozenten habe. Ich hörte seine Vorlesung "Einfach in die Sprachwissenschaft", da war er schon längst eremitiert. Das Buch habe ich weniger gelesen denn damit gearbeitet. Es ist vielleicht die einzige Einführung in die sehr komplexe Materie, in der alle indoeuropäischen Sprachen von einer rekonstruierten Grundsprache her verglichen und in ihren Entwicklungen dargestellt werden. Wobei es eigentlich nur die "alten" Sprachen sind, die ersten Vertreter ihrer Sprachzweige und -familien, von denen wir durch schriftliche Zeugnisse Kenntnis haben.
As a Classics major hoping to do postgraduate study in comparative Indo-European linguistics, I was in need of a basic introduction to the field, and this work of Szemerenyi seemed useful. However, INTRODUCTION TO INDO-EUROPEAN LINGUISTICS (a 1990 translation from the German by the author himself) was something of a disappointment.
The datedness of the theory in the book heavily reduces its usefulness. While some recent work has been taken into account--it explains, for example, that the Praenestine brooch is a forgery--Szemerenyi was the last holdout against laryngeal theory, perhaps the most useful breakthrough in PIE phonology of the 20th century. He does explain the theory (or a strawman), but calls it highly incredulous and doesn't use laryngeals in reconstructions of vocabulary. Therefore, the view of the proto-language's phonology in this book follows the beliefs of over 50 years ago, and will not match anything in the other contemporary handbooks. While it is important to understand the various views that have been held over the last two hundred years, one feels cheated if one has placed a great deal of trust in, for example, the existence of two schwas when the laryngeal theory removes the need for such sounds.
The book's saving grace is that it does have a copious biography, which makes it somewhat useful. However, other current handbooks usually have the same amount of useful references. If you are interested in comparative Indo-European linguistics, I would recommend any other handbook published in the last 15 years. Lehmann's THEORETICAL BASES OF INDO-EUROPEAN LINGUISTICS is, in my opinion, the best and most friendly for people entirely new to the field.