Brian Stubbs, a well-respected linguist with numerous publications on the history of Uto-Aztecan (UA) languages under his belt,¹ has finally released his magnum opus, a compendium of lexical, phonological, and grammatical data that provides evidence for infusions of ancient Near Eastern languages in Uto-Aztecan grammar and lexicon.

The claim for these infusions is based on the linguistic notion of cognate. Two words are cognate if it can be demonstrated that they both have a common historical source and that their sound (and meaning) differences are due to normally occurring linguistic change. For instance, the English words father and thin are cognate with Latin pater and tenuis. They do not look exactly alike, but the correspondences between the sounds of English and the sounds of Latin are regular and help establish these pairs of words as cognates.

The richest source of cognates is found in the basic vocabulary of a language: body parts, kin terms, natural phenomena, and so forth. Collecting enough cognate sets eventually yields regular sound correspondences that can be used to reconstruct the ancestor language and provide a rough timeline for the changes that have taken place in each daughter language. In his book, Stubbs presents over 1,500 cognate sets that show regular correspondences between Egyptian and Semitic on the one hand and Uto-Aztecan on the other.

Central to Stubbs’s proposal is the division of the Semitic influence into two varieties: “Semitic-kw” and “Semitic-p.” The second chapter presents cognate sets that demonstrate Semitic-kw contributions to Uto-Aztecan. Semitic-kw is so-called because of the correspondence between Semitic b and Uto-Aztecan *kw.² Consider the following examples of this correspondence (67–68):³

4 Hebrew bāšel ‘boiled’ ~ UA *kwasïC⁴ ‘cook, boil, ripen’

5 Hebrew bāśār ‘flesh, penis’ ~ UA *kwasiC ‘tail, penis, meat’

6 Hebrew blʕ / bālaʕ ‘swallow (v)’ ~ UA *kwïluC ‘swallow’

7 Hebrew bāmā (< *bahamat) ‘back, hill, mountain ridge, high place’ ~ UA *kwahama ‘back’

In each of these sets, Semitic *b corresponds to Uto-Aztecan *kw. Other correspondences found in Semitic-kw include Uto-Aztecan *(h)o, *(h)u, and *w for Semitic gutturals (χ, ʁ, ħ, ʕ), and UA *ts for Semitic ṣ and ṭ. This chapter represents the oldest stratum of Stubbs’s research. He presented a summary of his initial findings in a FARMS report;⁵ the data in the present work does not differ in their essentials from the earlier summary, and the cognate sets still hold up after all these years.

Chapter 3 discusses the pronouns of Uto-Aztecan. In this chapter, Stubbs tries to make the case for an infusion of pronouns from Semitic and Egyptian into Uto-Aztecan. Since pronouns are typically little words or even affixes, there is a far greater likelihood that chance resemblances will show up. The following table shows Proto-Uto-Aztecan pronouns:⁶

	singular	plural
1st person	*(i-)nï	*(i-)ta(-mï)
2nd person	*ï(-mï)	*ï-mï
3rd human	*pï	*pï-mï
3rd non-human	*a	*a-mï

The plural forms consist of a stem followed by a plural suffix (Stubbs argues that the Uto-Aztecan plural suffix is actually *-ima, which is a nice match for the early Hebrew masculine plural suffix *-īma). Removing the suffix (and the prefixed *i-) leaves behind single syllable forms. However, the chances that any two languages will show similarities in such small formatives is fairly high. (It is something of a parlor trick among linguists to find false cognates between any two arbitrarily chosen languages; it is surprisingly easy.) What is needed, then, is to demonstrate that there is a constellation of corresponding forms between the two languages that share form and meaning. While most of the examples in this chapter are easily disputed, Stubbs hits the jackpot with the correspondences between Semitic imperfective prefixes and Classical Nahuatl pronouns (86, reproduced below):

	Hebrew/Semitic sing.	Hebrew/Semitic plur.	Maghrib Arabic	Nahuatl
1st	ʔe-/ʔa- ‘I VERB’	ni-/na- ‘we VERB’	n- ‘I VERB’	neʔ-wa ‘I’
2nd	ti-/ta- ‘you sg VERB’	ti-/ta- ‘you pl VERB’	t- ‘you VERB’	teʔ-wa ‘you sg’
3rd	yi-/ya- ‘he VERBS’	yi-/ya- ‘they VERB’	y- ‘he VERBS’	yeʔ-wa ‘he’

For there to be occasional coincidences in form and meaning is expected. To have a whole array of pronouns agree so closely is much more convincing and strengthens the argument considerably.

Chapter 4 presents Egyptian contributions to Uto-Aztecan. There are some interesting grammatical parallels that Stubbs explores in this chapter. Chief among them is the correspondence between the Egyptian perfective/stative suffix -i and Uto-Aztecan *-i, which is final on intransitive, passive, or stative verbs. Other passive and stative markers are shown to correspond as well (87–88).

Sound correspondences between Egyptian and Uto-Aztecan differ somewhat from those of Semitic-kw and Uto-Aztecan. For instance, Egyptian b corresponds with UA *b or *p rather than with *kw (93–94):

137 Egyptian bbyt ‘region of throat’ ~ UA *papi ‘larynx, throat, voice’

138 Egyptian bšw ‘spittle, vomitus’ ~ UA *piso- ‘vomit’

139 Egyptian bnty ‘breasts’ ~ UA *piCti ‘breast’

Chapter 5 is by far the longest chapter of the book; in it Stubbs presents almost one thousand cognate sets demonstrating Semitic-p contributions to Uto-Aztecan. The sound correspondences that are attributed to Semitic-p are the same as those attributed to Egyptian. This suggests that speakers of both Egyptian and Semitic-p came into contact with Uto-Aztecan speakers at about the same time and that the Semitic-kw infusion represents a different contact situation or contact at a different point in time.

To Latter-day Saints, a scenario immediately presents itself to explain two separate Semitic infusions, but Stubbs is careful to avoid this sort of speculation and to let the data speak for itself. As with Semitic-kw, the correspondences between Semitic-p and Uto-Aztecan are regular, and the sheer mass of cognate sets is overwhelming. Here’s a short sampling (158–93):

527 Hebrew bārāq ‘lightning’ ~ UA *pïrok ‘lightning’

528 Hebrew béged/bāged ‘garment, covering, clothing’ ~ UA *pakati ‘shirt’

534 Hebrew batt ‘daughter’ ~ UA *pattï ‘daughter’

569 Hebrew rʔw/rāʔā ‘see’ ~ UA *tïwa ‘find, see’

631 Aramaic ħamar, Hebrew ħɛmɛr ‘wine’ ~ UA *kamaC ‘drunk’

711 Hebrew kɛlɛb, kalb- ‘dog’ ~ UA *kalop ‘fox’

Chapters 6 through 8 treat various comparative matters, including how this proposal solves several outstanding problems in Uto-Aztecan historical phonology and some speculation concerning the actual language represented by Semitic-p. Chapter 9 is a brief conclusion. There are four appendices, consisting of a summary of the sound correspondences, an English index to the cognate sets, and Hebrew and Egyptian indices.

This book is intended for linguists, Semiticists, Egyptologists, and Uto-Aztecanists. Stubbs includes an introductory chapter providing basics of historical linguistics and short summaries of Semitic, Egyptian, and Uto-Aztecan languages intended to help nonspecialists get their bearings in what follows. The scholarship throughout is sound. Stubbs has a good track record of academic publication in Uto-Aztecan studies, and he is just as careful with his treatment of the present material as he is in his more traditional Uto-Aztecan work.

My greatest complaint is that this book did not go through the standard academic editorial and review process. On the first page, Stubbs states that Uto-Aztecanists, Semiticists, and Egyptologists probably will not be receptive to his proposal or take seriously the notion that Old and New World languages could have mixed in such a fashion. He may be right about his peers, which would make standard academic review more difficult. However, the editorial and review process have the benefit of helping authors explain themselves more effectively to those who disagree or do not understand. It is obvious that Stubbs understands perfectly well what he is saying; however, his book fails in many places to say it clearly and directly to others. I was always able to puzzle it out, but the data and the arguments are complicated, and peer review and skilled editorial assistance would have been helpful to readers.

At first glance, this book seems to fall in with the type of linguistic crackpottery that claims Hebrew (or Sanskrit) as the mother tongue for all of the world’s languages, or that purports to relate Basque to any number of disparate languages.⁷ The book is dense, self-published, and in sore need of careful editing—none of which immediately commends it to the serious reader. However, Stubbs has something the language eccentrics do not have: the training and experience, together with extensive accurate data, to back up his extraordinary claim of significant Old World linguistic influence in Uto-Aztecan, a New World language family. It is definitely worth the trouble to work through this book.

Dirk Elzinga

BYU Studies Publications