LINGUIST List 33.1101
Fri Mar 25 2022
Review: Greek, Modern; Linguistic Theories; Syntax: Givón (2021)
Editor for this issue: Amalia Robinson <amalialinguistlist.org>
Date: 10-Jan-2022
From: Ezra la Roi <ezralaroi
gmail.com>
Subject: The Life Cycle of Adpositions
E-mail this message to a friend Discuss this message Book announced at
https://linguistlist.org/issues/32/32-2629.html
AUTHOR: T. Givón
TITLE: The Life Cycle of Adpositions
PUBLISHER: John Benjamins
YEAR: 2021
REVIEWER: Ezra la Roi, Ghent University
SUMMARY
Thomas (Talmy) Givón’s latest book examines the life cycle of adpositions, focusing in particular on the tail-end of the life cycle when adpositions end up as verbal affixes. As one may expect from Givón’s track-record, his examination seeks to go beyond a linear diachronic analysis. He weighs the theoretical import of the changes observed in this life cycle (esp. the change from nominal to verbal domain) as well as the syntactic constraints (e.g. word order) and discourse-pragmatic processes (e.g. zeroing) which lie at the heart of the life cycle of adpositions. After a typological introduction to the syntax and diachrony of adpositions, he provides a corpus study of a few books of Homer’s Iliad, as its language, he argues, contains a synchronic distribution of diachronic variants which supports the life cycle of adpositions that he devises. The book closes with a corpus study of adpositions in the history of English which, due to its VO word order, provides the mirror image of the developments in Homeric Greek (with its OV word order).
Chapter 1 introduces the diachronic study of adpositions. Givón describes their functional domain as typically marking indirect object roles which express either the location of the subject, the source/goal of its motion, or the source/goal of the object’s motion. Next he discusses Greenbergian correlations of prepositions with “basic” VO order and postpositions with OV and goes on to point out the diachronic relevance of the cross-linguistic conditional association which states that languages with verb-attached adpositions will also have nominal attached verbs. Using a survey of three unrelated languages (i.e. English, Kinya-Rwanda (Bantu), and Rama (Chibchan)) he then hypothesizes how their originally nominal adpositions may have changed to verb affixes through various discourse-pragmatic mechanisms. He posits that English verbs with prefixes (derived from both Romance and Germanic sources, e.g. ‘con-tain’ vs ‘for-get’) are the result of an older cycle reflecting the older OV syntax of Latin and Germanic, while a more recent cycle reflecting the more modern VO syntax of English has caused prepositions to be stranded behind the verb (e.g. ‘she shut him up’). In Givón’s view, the discourse-pragmatic processes that can help shift nominal adpositions to the verb involve zeroing out of objects (indirect or direct) when they are predictable from their anaphoric or generic context or have higher topicality (cf. antipassive ‘she often reads in bed’ or anaphorically accessible ‘we were at the house when she came by [=by the house]’). On page 31 Givón presents a preliminary version of his life cycle and connects a syntactic-typological prediction. His life cycle consists of three major stages: in stage 1, adpositions are case-role markers attached to or associated with nominals; in stage 2, the nominal object is in the appropriate discourse context of anaphoric/generic predictability zeroed out and the adposition is detached or free-floating; in stage 3 the adposition re-attaches itself to the remaining lexical word, the verb. On the basis of his survey, Givón makes two diachronic predictions: (1) in VO languages (e.g. English), stranded adpositions are more likely to become post-verbal clitics and eventually verb suffixes, whereas (2) in OV languages (e.g. Homeric Greek, Proto-Germanic/Romance), stranded adpositions are more likely to become pre-verbal clitics and eventually verb prefixes. In the chapters that follow, he details how the synchronic variants in Homeric Greek can be reinterpreted as evidence for the life cycle of (2), whereas the last chapter of the book supports the life cycle for (1).
Chapter 2 provides the preliminaries to the analysis of Homeric Greek and the baseline of the life cycle, i.e. pre-nominal prepositions. Givón opens with three related claims which he forcefully puts forward: (1) based on just two grammars of Ancient Greek, i.e. Monro’s of Homeric Greek (1891) and the Cambridge Grammar of Classical Greek (2019), he claims that Classical Greek grammars fail to predict the syntactic effects of suffixal case, prepositions, and their combinations; (2) similarly, he claims that they do not offer syntactic or diachronic explanations of those cases where prepositions are stranded as they entertain the notion of tmesis (i.e. free standing adverbs through severance); and (3) he notes that verb forms with prefixes (= preverbs) are “augmented” much more often than those without, “which most Classicist grammarians have chosen to ignore” (p.36). He subsequently pits his theoretical approach against what he believes is the norm in Ancient Greek linguistics, as his method of Internal Reconstruction (i.e. inferring diachronic processes from synchronic evidence based on language universals and typological diversity) “diverges radically from the Classicist obsession with Comparative Reconstruction of Indo-European proto forms” (p.36). Thus, his approach focuses on synchronic variation of attested forms, their functions, and their respective quantitative textual distributions. Givón notes his caveats and apologia for phonological and grammatical transcription of Homeric Greek (e.g. letters, vowel elision, accents, word division, and status), the text, and the translation. The remainder of the chapter illustrates the usage of pre-nominal prepositions in books 1 and 2 of Homer’s Iliad; Givón observes that in those two books (out of 24) all pre-nominal prepositions are used to mark indirect objects (in a concrete or more abstract fashion) of the obligatory or optional kind. He finds a slight preference for OV order of preposition-marked indirect objects in book 1 of the Iliad (55%), which he takes as evidence for why prepositions cliticized pre-verbally in Homeric Greek, since, according to his reconstruction, OV order must have been even stronger before.
Chapter 3 focuses on the diachronic target, i.e. pre-verbal prepositions as prefixes. Givón gives four diachronically different syntactic realizations of the pre-verbal target constructions: (a) nominal indirect object overtly present, (b) not overtly present but accessible, (c) not present nor accessible but a transparent association between the preposition and the verb, and (d) semantic arbitrariness. The end of this cline consists of cases where the preverb is semantically arbitrary (cf. English ‘speak up’ or Iliad 1.87 ‘anaphaineis’ ‘you pronounce’, ‘ana-‘ meaning ‘up’). Using this diachronic trajectory he classifies the different preverbs as one of these stages on the basis of the semantic transparency and syntactic functioning of the preverb. A contrastive count of these preverbal prepositions with their previous stage as pre-nominal prepositions (Chapter 2) subsequently reveals that some prepositions still have a higher distribution as prenominal (e.g. ‘en-‘ ‘in’, ‘ekh/k-’ ‘out’, ‘eis-’ ‘to’, ‘peri-‘ ‘around’) whereas others already have a majority preverbal use (e.g. ‘apo-’ ‘from’ or ‘ana-’ ‘up’). These figures are said to reveal a diachronic continuum toward the diachronic target, something which he supports with percentages illustrating that stages (c) and (d) are in the minority (resp. 14.7% and 20.4%) as opposed to stages (a) (50.9%) and (b) (14%).
Chapter 4 details the stage preceding cliticization as preverb, i.e. detached (“severed”) prepositions in Homeric Greek. As expected, these detached prepositions precede the verb (except in some occasions labeled problematic but left unexplained, pp. 94-96). Given that detached prepositions develop into preverbs, Givón argues that their indirect objects are zeroed out in different stages, first losing the syntactic restriction and then the semantic restriction to the indirect object. He supports these stages with examples of different kinds (e.g. examples where a zeroed indirect object is anaphorically accessible or where it is not accessible anymore). The relative textual distributions in two books of the Iliad are again adduced to support these stages (e.g. 19.5% in stages (c) an (d) versus 80.5% in stages (a) and (b)). Another point of attention is the role played by the vast group of second-position “particles” which can follow a detached preposition. Givón characterizes these particles as a form of clause chaining. He then tentatively hypothesizes that all these particles could reflect an earlier use of detached propositions in some context of contrast or emphasis, since they (on the whole) serve these functions. The higher distribution of these occurrences with a particle (72.3%) would in his view support his diachronic reconstruction.
Chapter 5 presents an alternative solution to the vexed issue of the unpredictable “augment” in Homeric Greek, which in Classical Greek functions as a past tense marker for verbs in the indicative. Givón claims that the Homeric augment should be reinterpreted as an earlier cycle of pre-verbal prepositions, viz. a phonologically reduced variant of ‘en-’ ‘in’, ‘ekh/k-’ ‘out’ or ‘eis-’ ‘to’. Contrary to the prevailing opinion among linguists of Ancient Greek, Givón therefore does not limit the augment to past indicative verbs, despite its target in Classical Greek; rather, he claims that augments can even be found with participles, presents, or subjunctives. Since the augment has what he calls a “mixed bag of conditioning contexts” (p. 110) in Homeric Greek, Givón claims that all previous explanations fail to fully explain the distribution of the augment. However, he observes “that the verbal-clausal contexts in which the ‘Augment’ e- is found are highly predictable, with the vast majority involving clauses with verbs of location, motion or direction that, in most languages, take adposition-marked indirect objects” (p. 113). The main part of the chapter describes the verb types and syntactic contexts in which the augment is found in book 1 of the Iliad: (1) intransitive verbs with indirect object or with it being zeroed out (e.g. ‘be in/at’, ‘come to’, ‘look at’, ‘hear from’), (2) bi-transitive verbs with indirect object or with it being zeroed out (e.g. ‘bring x to/from’, ‘put x at/in/out’, ‘tell x to’, ‘receive x from’), (3) residue of “plausible diachronic derivatives” (p. 151) of verbs from the previous categories. These contexts make up 71.6% of the instances in book 1 of the Iliad. The fact that the instances of type 1 and 2 are found with various adpositions in the clause (e.g. pre-nominal preposition, post-nominal post-position, or detached preposition) is reinterpreted by Givón as diachronic evidence that these instances recapitulate or augment the original sense which has fossilized as augment. The residue class of 28.4% (e.g. with volition verbs, manipulative verbs, ‘know/see’ and ‘be in/at’ verbs) is explained on typological grounds, that is, there are languages which code their objects with adpositions, e.g. ‘I want to eat’, ‘I told him to leave’, ‘be in’. After his survey of the evidence (drawn only from book 1 of the Iliad), Givón concludes that his predictions are preferable to those offered in the literature, albeit without falsifying those predictions.
In the next chapter, Givón takes his hypothesis about the augment a step further. Most importantly, he argues that those verbs which have an augment but are preceded by a preverb are also mostly found in the same syntactic contexts. He suggests that, as a synchronic variant, these contexts provide diachronic evidence that those preverbs were cliticized before the augment as recapitulation of the original adpositional sense of the fossilized augment. He then illustrates the similarity in syntactic context of these variants with various preverbs (e.g. ‘kath(a)-’ ‘down’, ‘an(a)-’ ‘up’, ‘par(a)-’ ‘by’, ‘ep(i)-’ ‘to’, ‘met(a)-’ ‘with’) with examples from the first 3 books of the Iliad (making up 89.2% of the total amount), while suggesting that these preverbs recapitulate or augment the senses of ‘en-’ ‘in’, ‘ex/k-’ ‘out’, ‘eis-’ ‘to’ from which he thinks the augment has developed. Finally, he justifies his hypothesis in scientific terms of falsification and claims that only his hypothesis can serve as a “plausible etymology” (p. 165) to explain why it is ‘e-’ that is prefixed to these verbs.
The last chapter of the book provides us with a mirror image of the developments that were discussed for Homeric Greek and its prehistory, in the form of a diachronic survey of the life cycle of adpositions in the history of English. Using small case-studies of different diachronic varieties of English (from 14th-century Chaucer to modern-day English), he demonstrates that adpositions developed in the reverse direction in English, as pre-nominal prepositions developing into post-verbal stranded prepositions which eventually turn into post-verbal clitics. As a form of closure, Givón points to the theoretical import of his studies, the most important being that synchronic syntactic variants reveal a diachronic continuum and even Homeric Greek, which is considered a unique language by many, reveals language universals in diachrony like other typologically diverse languages.
EVALUATION
Although theoretically interesting and also stimulating in terms of some findings, Givón’s book is probably best evaluated as a “mixed bag” itself. The book’s merits surely lie in the internal syntactic reconstruction of the data, which offers insightful views on how adpositions develop diachronically and which discourse-pragmatic processes enable their evolution. As such, chapters 2 through 4 on Homeric Greek and Chapter 7 on English surely yield some relevant insights, both about the distribution of adpositions and preverbs and their syntactic or pragmatic motivation. That being said, there are some shortcomings and problematic aspects to the book which I am afraid could impede the use of the book by future scholars.
The most critical shortcoming is the fact that, as evidenced by Givón’s bibliography of just three pages for a 200-page book, there is no satisfactory effort to engage with the existing literature on adpositions, neither of general linguistics nor specifically of Ancient Greek or English linguistics. For the general linguistic audience, Givón does not engage at all with books on adpositions (e.g. Hagège 2010 or Kurzon & Adler 2008) nor with important earlier findings on the diachrony of adpositions (see the discussion of adpositions by many chapters in Narrog & Heine 2011 and by Plank 2011, who also discusses the role of diachrony and universals). Despite the promise of “well deserved homage to the traditional Classical scholarship” on the book’s cover, the book’s engagement with the audience of Ancient Greek linguists disappoints as well. On the one hand, their depiction as obsessed with comparative reconstruction (p. 36) is both unfair and untrue and, on the other hand, their hypotheses on adpositions and the augment are not tested. Obviously, not all linguists working on Ancient Greek depend on the comparative method, as this method is typically used by Indo-Europeanists, while others have rightly analyzed the corpus data from a variety of theoretical perspectives, such as discourse-pragmatic, syntactic, and cognitive frameworks (cf. Logozzo & Poccetti 2017 and Mocciaro & Short 2019). What Givón presents as new, then, is not new at all: (i) as early as Kühner-Gerth (1898: 449-450 & 530-531), still one of the most highly praised standard grammars of Ancient Greek, it has been said that what is called ‘tmesis’ is actually a different diachronic variant; (ii) Ancient Greek linguists know very well that the augment had not yet stabilized as a rule in Homeric Greek (as it stabilized as a past marker only later on in Classical Greek), but was rather ruled partially by metrical needs and pragmatic factors of usage; and (iii) Indo-Europeanists have identified and analyzed the life cycle of adpositions to preverbs as well (see Hewson and Bubenik 2006 and Fortson 2010: 154-156) but are left out of Givón’s discussion. Although I am no expert on English, similar remarks probably apply there as well, as the only reference given in the discussion of the corpus evidence for English is to Givón’s own work.
In addition, there are several problematic aspects to this book with regard to how it develops its hypotheses. Firstly, the use of limited amounts of corpus data undermines Givón’s analyses, especially where he bases his theory that the augment is an old preposition on quantitative textual distributions of just one book of the Iliad (i.e. 611 lines) although it consists of 24 books (i.e. over 15.000 lines). In fact, even for the data from that one book his explanation is not absolute nor would it explain obvious counterexamples of an augmented past verb where an (optional or obligatory) indirect object cannot be added in the interpretation (e.g. ‘e-trephon se’ ‘I reared you’ Il. 1.414). Secondly, his hypothesis that the ‘e- augment’ is an old preposition (‘en-’ ‘in’, ‘ekh/k-’ ‘out’ or ‘eis-’ ‘to’) is inconsistent because it is sometimes realized as a long vowel and other times it is not (‘ēlthon’ ‘I came’ vs ‘elthön’ ‘having come (participle)), and crucially contradicts the less evolved status of these alleged preposition sources which are still typically pre-nominal in Homeric Greek (see p. 81). Givón’s theory also fails to explain why the augment in these fossilized prepositions would have developed into a past tense marker in Classical Greek. Thirdly, his attempt to use Homeric Greek as a natural language causes problems for his analysis, as metrical alternations for long vowel or short vowel augmentation in the same verb are not explained, see Il. 1.399 ‘ēthelon’ vs. 4.380 ‘ethelon’ ‘they wanted’ (cf. Fortson 2010: 154-157 on the artificial nature of Homeric Greek). Fourthly, the Greek is incorrectly glossed on many occasions, e.g ‘ēke’ ‘she sent’ from ‘iēmi’ ‘to send’ is glossed as ē-k-e to-send-3s, which disregards the long vowel in the stem (see p. 93 and p. 99), the ‘-ai’ and ‘-sthai’ infinitive ending variants are glossed as irrealis forms (e.g. p. 119), a subjunctive is glossed as a future (p. 123 ex. 53) and some main verbs as non-finite forms (e.g. the pluperfect ‘eilelouthas’ ‘you came’ or ‘ekluon’ ‘they listened’ on p. 122 ex. 47 and ex. 48). Similarly, Givón’s argument that intervening second-position particles (in Wackernagel position) show the syntactic status of detached prepositions is problematic, since these particles could not have been in the first position of the clause. Finally, one could argue that his augment hypothesis is to some degree circular, because (i) the interpretation of a zeroed indirect object (either obligatory or optional) is possible in principle in most syntactic clause types (and even facilitated when there is already a preverb before the ‘augment’ see chapter 6), (ii) the augment is anachronistically interpreted with adpositional functions, something their source does not have in Homeric Greek e.g. ‘eklue’ ‘s/he listened to’, where e- is interpreted as ‘eis-’ ‘to’ by Givón but ‘eis-’ ‘to’ had no such functions in Homeric Greek (see Luraghi 2003: 107-117), and (iii) a typological parallel should not provide the sole basis for a diachronic reconstruction, e.g. ‘want to’ for ‘e-thelon’ ‘I/they wanted’, whose object was never marked by an adposition in Ancient Greek, especially not ‘eis’ ‘to’.
Despite these shortcomings, I would still recommend the book both to general linguists and to those working on Ancient Greek or Indo-European, provided its findings are approached with caution. Disregarding the chapters on the augment and the lack of engagement with the literature, Givón presents findings which challenge ideas on adpositions and their diachrony and deepen our understanding of adpositions and preverbs.
REFERENCES
Fortson, Benjamin W. IV. 2010. Indo-european language and culture : an introduction. 2nd ed. Chichester: Wiley-Blackwell.
Hagège, Claude. 2010. Adpositions. Oxford : Oxford University Press.
Hewson, John & Vít Bubeník. 2006. From case to adposition : the development of configurational syntax in Indo-European languages. Amsterdam ; Philadelphia : Benjamins Pub. Co.
Kühner, Raphael & Bernard Gerth. 1898. Ausführliche Grammatik der griechischen Sprache. Zweiter Teil: Satzlehre. Erster Band. Hannover : Hahnsche Buchhandlung.
Kurzon, Dennis & Silvia Adler (eds.). 2008. Adpositions: pragmatic, semantic and syntactic
perspectives. Amsterdam ; Philadelphia: John Benjamins Pub. Company.
Logozzo, Felicia & Paolo Poccetti (eds.). 2017. Ancient Greek Linguistics: New Approaches,
Insights, Perspectives. Berlin; Boston: de Gruyter.
Luraghi, Silvia. 2003. On the meaning of prepositions and cases : the expression of semantic roles in
Ancient Greek. Amsterdam : Benjamins.
Mocciaro, Egle & William Short (eds.). 2019. Toward a Cognitive Classical Linguistics: The Embodied
Basis of Constructions in Greek and Latin, Warsaw, Poland: De Gruyter Open Poland.
Narrog, Heiko & Bernd Heine. 2011. The Oxford handbook of grammaticalization. Oxford : Oxford
university press.
Plank, Frans. 2011. Where’s diachrony? Linguistic Typology 15(2). doi:10.1515/lity.2011.030
ABOUT THE REVIEWER
Ezra la Roi is currently a PhD student in the linguistics department at Ghent University in Belgium. In his research he examines the history of counterfactuals from Archaic to Post-Classical Greek (VIII BC- III AD) whilst testing diachronic typologies of counterfactuals. In his work he combines the most recent insights from theoretical linguistics (esp. linguistic typology, historical linguistics and pragmatics) with rigorous corpus-based analyses of various stages of Ancient Greek. His recent publications concern modality, counterfactuals, insubordination, habituals and the category change of mood forms.
Page Updated: 25-Mar-2022