More than two thousand years ago, the complex grammar of the ancient Persian language was drastically simplified within five generations, coinciding precisely with the period when the kings of the Persian people ruled the Achaemenian Empire – the largest empire ever created until that time, reaching from southern Egypt to Central Asia, from the shores of southeastern Europe to the Indus.Footnote ¹ This book explains that correlation: how Persian imperialism two and a half thousand years ago, unprecedented in scale, triggered sudden grammatical effects that are still evident in the modern Persian language today. The Persian language was morphologically reduced and restructured as a regular and even predictable result of the demographic upheaval created by the Persian Empire itself. This insight prompts, in turn, meaningful revisions to the currently dominant historical narrative about ancient Persian imperial society and the ruling class of this paradigmatic empire of antiquity. The new patterns of relationships instigated by the mighty empire fostered an unplanned domestic imperial melting pot hitherto scarcely detected and understood by modern historians.

Old Persian, Middle Persian, New Persian

Specifically, this book investigates the first of the two transitions in the history of Persian, from the oldest attested stage of Persian, called Old Persian, to the subsequent stage, called Middle Persian. A word about these designations is in order. Scholars have a convention for designating stages of evolution in the history of a language by the terms Old, Middle, and New. We have, for example, the Old English of Beowulf, the Middle English of Chaucer’s Canterbury Tales, and the New English still used today. The separate stages are, to a degree, artificial categories in that each stage is changing and dynamic in itself. There is dialectal, chronological, and regional variation within each stage. Shakespeare’s New English is not the same as my own New English, nor is my American New English quite the same as that of my colleagues from London. Within London itself one can hear different varieties of English. Although these chronological distinctions are therefore conventional and admit much variation, linguists have sound criteria, based primarily on the grammar of English, for the division of such broad historical phases of the language. The grammar of the language in each of these three different stages is distinctly different enough from that of the others that they each can be described by separate synchronic grammars.Footnote ²

The same convention – Old, Middle, New – is used for the historical stages of the Persian language, a usage nearly two hundred years old.Footnote ³ Old Persian is attested in the royal inscriptions of the Achaemenid Persian dynasty, which ruled their extensive empire from the sixth to the fourth centuries bce. It was written in a simple cuneiform script specially designed to express it. Middle Persian, attested in inscriptions from the first century bce onward, was used by the Persian kings of the Sasanid dynasty from the third to the seventh centuries ce and, until today, as a literary language of the Zoroastrian religion.Footnote ⁴ It was written in a much-adapted variety of the Aramaic script. The third stage, New Persian, first appears in records of the ninth and tenth centuries in Central Asia after the Muslim colonization of that region. It was written predominantly, though not exclusively, in a variety of the Arabic script, as it still is today, and it is spoken today in mutually intelligible varieties in the Islamic Republic of Iran, the Islamic Emirate of Afghanistan, and the Republic of Tajikistan (where it is written in the Cyrillic script). As with the conventional terms for the stages of English, each designated historical stage of Persian is artificial to the extent that it exhibits dialectal, chronological, and regional variations, but there are clear criteria for dividing each stage from the others. Middle Persian is different from Old Persian in specific, essential features of its grammar, and so is New Persian from Middle Persian. It has long been evident to scholars that Old Persian became Middle Persian or something close to it – using these terms in the general, imperfectly analytical ways just explained – already by the time Alexander of Macedon launched his invasion of the Persian Empire in 334 bce. This book explains why the Persian language changed so rapidly and permanently by that time.

Hypotheses about the “Simple” Persian Language

The Persian language today is renowned as having a “simple” grammar. That is, when compared to other formerly or currently widespread languages, such as Arabic or Russian or ancient Greek, it has a simple morphology. This means it is minimally inflected: there are few changeable forms of words to express grammatical relations between words to make cogent sentences. This makes it easier for adults to learn.Footnote ⁵ In respect to its inflectional simplicity, Persian today is like English, a comparison often casually drawn over recent centuries. For example, the British orientalist William Jones (1746–1794) wrote in his widely read grammar of Persian for English-speakers, “The reader will soon perceive with pleasure a great resemblance between the Persian and English languages, in the facility and simplicity of their form and construction.”Footnote ⁶ The Swiss philologist Franz Misteli (1841–1903) compared English and Persian grammar closely in his volume of 1893 on linguistic typology.Footnote ⁷ He held that “historical development” of a nation changes its language, and the more removed a people is from engaging in “history,” the more its language remains the same. He posited that Persian and English both similarly lost the ancient inflection of their common ancestor due to parallel circumstances (which he did not specify further).Footnote ⁸ Edwin Lee Johnson (1874–1947), author of an Old Persian grammar, wrote that “The relation of the New to the Ancient Persian presents something of a parallel to the relation of Modern English to Anglo-Saxon, in that an analytical language has been developed from one highly inflectional.”Footnote ⁹ Even without a comparison to modern English, the relative simplicity of New Persian grammar was noticed by many others, too. Meerza Mohammad Ibraheem (circa 1800–1857), Professor of Arabic and Persian Languages at the East India College near London, wrote in 1841 that Persian grammar was “easy and simple.”Footnote ¹⁰ Three hundred years earlier, the blind Egyptian Arab physician Dāwūd of Antioch (d. 1599) wanted, as a young man, to learn Persian; his foreign teacher told him that Persian “is easy for anybody, my son,” and he offered to teach him rare knowledge of ancient Greek, which hardly anybody there knew, instead.Footnote ¹¹ The Prussian philosopher Friedrich Engels (1820–1895), an educated nonspecialist, wrote to Karl Marx that the acquisition of Persian grammar was “child’s play” compared to the grammar of Arabic.Footnote ¹²

What is not widely known today is that Persian acquired most of its relative simplicity of inflection more than two thousand years ago, already at the stage of Middle Persian. As Walter Henning (1908–1967), a preeminent specialist in Iranic languages, noted in 1958, “With respect to morphology, ‘Proto-Iranic’ was comparable to Sanskrit or Greek, whereas Western Middle Iranic [including Middle Persian] is on a par with English.”Footnote ¹³ But when the Persian language was first written in ancient inscriptions of the late sixth century bce, its morphology was still quite complex. The Old Persian of that early time still had a highly inflected grammar with very many unpredictable word forms that learners needed to use to express themselves competently with native speakers. While not quite as ornately inflected as Old Avestan – an archaic Iranic cousin preserved only in a small, fixed corpus of ancient liturgical texts, maintained over centuries by daily priestly recitation – Old Persian was indeed more like ancient Greek or Sanskrit in its morphology. Middle Persian, by contrast, while not as simple in its inflections as New Persian, was already nearly so, as has been observed by Persian specialists.Footnote ¹⁴

It may seem bold to claim that the creation of a vast empire triggered the restructuring of the very grammar of the language of its ruling elite. In fact, the claim, which this book verifies, was first articulated with respect to Persian by others as hints or suggestions, but it has never been systematically explained and defended, nor have the extraordinary implications of the claim been drawn out. Persian and Iranic language specialists such as Meillet (1866–1936),Footnote ¹⁵ Szemerényi (1913–1996),Footnote ¹⁶ and UtasFootnote ¹⁷ have intuited a connection between population contact and the reduction of Persian morphology, but they neither had the linguistic models available now nor did they provide sufficient evidence to support their tentative suggestions. The tendency in Persian studies generally, moreover, has been to focus on morphological simplicity in the evolution of New Persian, not Old or Middle Persian. The one concerted case to explain the reduction of Persian inflectional morphology in antiquity was made by a nonspecialist in Persian. Linguist John McWhorter’s book Language Interrupted (Reference McWhorter2007), on grammatical reduction through mass nonnative acquisition, has a chapter on Persian, along with other chapters diagnosing the problem for other languages that have undergone similar grammatical restructuring, including English, Mandarin, and Malay.Footnote ¹⁸ That book is the most important antecedent of the present one. Although I differ with his linguistic theory in some important details, I argue in what follows that McWhorter’s insight about the history of Persian is basically correct. Nevertheless, his chapter on Persian does not provide convincing historical contextualization for his arguments using primary sources, a minimum requirement for its use by historians. He likewise does not isolate the transitions from Old to Middle to New Persian clearly enough. To be fair, he addresses his fellow linguists who have largely synchronic concerns, not ancient historians or philologists, and his argument aims at general linguistic phenomena applicable to modern languages, for which Persian, like English, is merely one important example. McWhorter’s tentative historical-contextual explanations about Persian are speculative, based partly on outdated surveys of Achaemenian history rather than primary sources, and are in some respects clearly mistaken. Yet McWhorter made a breakthrough, neglected by both linguists and historians, which deserves recognition and further development.Footnote ¹⁹

What is required to explain this formative period in the linguistic history of Persian for historians and philologists is not just to explain the linguistic-historical model of analysis employed here, but also to combine relevant primary sources with that model, to demonstrate the relationship of the requisite social conditions with the grammatical change connected with the linguistic model. It is not so simple as saying that the Persian language “became imperial,” as Meillet intuited. The cause and the explanation are both much more complicated than that. The outcome here is that the application of new linguistic research demonstrates social facts about the ruling class of one of the most important ancient empires. This book bridges ancient history and linguistics in a new way beneficial to both areas of research.

Linguistic History and Historical Linguistics

Since its nineteenth-century inception, it is the field of historical linguistics that has largely controlled discussions of language change over long periods. Historical linguistic research is built on the extraordinary discovery that when languages change, they change according to regular patterns that can be expressed as systematic rules, sometimes even construed as laws of a language’s development. These are especially discernible in phonology, the system of a language’s sounds.Footnote ²⁰ This discovery enabled scholars to establish the relationships of languages to one another through clinching demonstration, to elucidate the histories of words systematically, to infer prehistoric cultural facts with high confidence, and to describe scientifically change in languages over time.Footnote ²¹ It prepared the way for the modern understanding of the nature of language itself and for the scholarly trend of structuralism and its aftermath, effects far beyond linguistics.Footnote ²² The valid and time-tested methods of historical linguistics have offered and continue to offer sensational discoveries. But historical linguistics is founded on description, not explanation with respect to human context. At its core, historical linguistics entails describing two historical stages of a language; the systematic changes required for the prior stage to reach the latter stage are discerned and defined through the accumulation of examples and counterexamples; then these changes are named and described with special terminology as regular processes. The results can be tested and refined. The comparative method of historical linguistics, as successful as it is, does not tell us much, however, about why languages change as they do and when they do or how individuals and groups of speakers came to speak differently. It does not address the diffusion of sound changes through a population of speakers of a language who adopt those changes. It cannot address the social context of language change, just as it cannot easily address the physiology of voice production. Other methods are required for such questions. Despite some exceptions, conventional historical linguistics, by its methods, must largely treat languages as objects apart from society.Footnote ²³ It is historians who are best able to narrate that social context. That said, the comparative method of historical linguistics does what it is supposed to do very well.

Although it is informed by historical linguistics from beginning to end, this book is not a work of historical linguistics in the sense just described. “We have to forget what it means to be historical linguists and become instead linguistic historians,”Footnote ²⁴ as Terry Crowley (1953–2005) remarked in his careful history of the genesis of Bislama, a modern creole derived from English that became the national language of the Pacific Island nation of Vanuatu. The present book represents a project somewhat more like that one. Without forgetting historical linguistics, it comprises linguistic history. In no way is this intended as a corrective to the fundamentally sound methods of comparative historical linguistics, which I wholeheartedly endorse and continue to teach to my own students. Rather, it is complementary with those methods. It also participates in fields of linguistics concerned with language as a social medium in that it is premised on the “conviction” articulated by Thomason and Kaufman “that the history of a language is a function of the history of its speakers, and not an independent phenomenon that can be thoroughly studied without reference to the social context in which it is embedded.”Footnote ²⁵ As Croft puts it, “Languages don’t change; people change language through their actions.”Footnote ²⁶ Although historical linguistic research provides necessary components of the present argument, it is already well established – many obscure details notwithstanding – that Old Persian and Middle Persian are different, how they are different, and which sounds and grammatical systems changed to which in the transition. I am therefore not outlining yet again how Old Persian and Middle Persian are formally different in a comprehensive way, nor do I seek to provide detailed grammatical descriptions of those differences. Such works, including grammars of the ancient stages of Persian, already exist in abundance and are mentioned throughout the notes and bibliography.Footnote ²⁷ I have chosen not to include extensive comprehensive charts to illustrate Persian morphology at different stages of the language’s history for this reason. I do not offer an explanation here in terms of language structure, such as we have already. My goal here is rather to explain why ancient Persian became so strikingly different from its immediate Old Persian antecedent and from its contemporary Indo-European relatives. The answer I propose is functional, meaning that it examines language change from the point of view of people’s efforts at communication. It has more to do with demography and social contact between groups speaking different languages than anything else. This necessitates not a comprehensive outline of ancient Persian phonology and morphology, but rather employing tested models correlating language change with patterns of population contact. The latter is primarily a historical, not a linguistic, topic.

The metaphor of a novel may work here. One can analyze the structure and contents of a novel and their variations over different editions without knowing much about the author’s life or the readers of the novel. Such an analysis is valid within those constraints. That said, understanding the life and experiences of the author, and the historical context of a novel’s authorship and readership and dissemination and reception will augment one’s understanding of the novel in other ways that do not necessarily invalidate a structural analysis. If the ancient Persian language were a novel, this book analyses aspects of the lives of its authors that helped determine the novel’s form.

More about Linguistic History

“Linguistic history” is not a rare expression in scholarship, but only limited attempts have been made to carve out a distinct domain for it as contrasted with historical linguistics, the latter of which is defined by its own rigorous methods.Footnote ²⁸ A survey of books with “linguistic history” in their titles shows that they do quite different things. Although some studies and books already bear the designation of linguistic history, the term has seldom been clarified. I venture here to define linguistic history in a way that may be useful to others who undertake similar projects. I also wish to distinguish it from different but closely related kinds of investigation. Readers who want to get to the matter of this book without lingering on fine methodological distinctions may skip ahead.

One might think linguistic history could simply be the history of a language – but that is quite insufficient. Languages exist solely as learned systems of communication between people: adults and children, individuals and groups. Language is not an abstract entity possessing disembodied independence. Unlike speakers, languages are objects of history without agency, not subjects. Therefore, linguistic history must be the history of languages told through the history of their speakers as far as their languages are concerned. It cannot treat languages effectively apart from speakers, nor can it ignore the grammatical and lexical facts of their languages. Just as cultural history is impossible without the history of people interacting with the world around them, linguistic history is impossible without the history of speakers and writers.

This has consequences. For example, it is not valid, from the strictest point of view of linguistic history, to speak of the “spread” of a language, as historians and others often do casually. Rather than “spread,” what occurs is the adoption of a language by new speakers, children or adults, and sometimes the shift of a population from one language to another via a stage of bilingualism. Likewise, it is not languages that move into new countries. A language’s speakers are the ones who move. Persian did not “spread,” but ancient Persians conquered others who had to interact with them, so some of them then learned Persian. This distinction may not matter in casual narration, but it matters greatly in linguistic-historical analysis. Furthermore, languages do not interact with each other and come into contact as autonomous entities; rather, people may use more than one language in the same life, or in the same situation, often leading to circumstances in which features are transferred from one language to another by bilingual individuals. In linguistic history people do things; languages do not. This must be a fundamental criterion of linguistic history, and its observance will help to avoid serious mistakes.

The foregoing, along with the rest of this book, may be enough to convey my intention with the expression linguistic history, but there are useful specific distinctions to be made with necessarily related fields. Linguistic history, because it concerns languages, must be informed by linguistics whenever that is relevant and possible. Linguistic history is especially closely connected with three kinds of linguistics, as the subfields are now divided: historical linguistics, sociolinguistics, and contact linguistics. It deals with some of the same problems as each of these three fields, but it is also distinct from them in its means and aims.

As already discussed, historical linguistics accounts for regular patterns of language change, mostly descriptively, and its primary tool has been the comparative method. For this reason, it has also been called comparative linguistics. Historical linguistics is diachronic by definition, producing an outline narrative of changes in the constitution and features of individual languages and reconstructed protolanguages. Because its explanations are structural, historical linguistics does not inherently need an account of the history of the speakers of a language. Normally, historical linguistics takes the grammars and lexica of languages and their components, and not their speakers, as its objects. Remarkably, the methods of historical linguistics work very well that way for its own purposes, but the relationship between social change and grammatical change – one of the subjects of this book – is thereby neglected.

Sociolinguistics, a newer field, attends to human relationships and society and their consequences for language.Footnote ²⁹ It arose partly in reaction to the explanatory limitations of historical linguistics just described. In practice, sociolinguistics is more typically concerned with behaviors of speech, and directly observable social linguistic behaviors especially. For example, it studies language variation conditioned by social factors such as class, gender, age, place of residence, attitudes toward specific languages, or the linguistic expression of solidarity with reference groups through pronunciation and usage. Such phenomena can be studied live, in process, documented in recordings and discovered through interviews. To focus on the study of social variation and social cues and signification in living speech is to emphasize current phenomena in a relatively short scale of time, ignoring bygone societies. The methodological emphasis on directly observable social linguistic behavior explains why sociolinguistics tends not to be extensively diachronic, although any study of change is inherently historical at least on a short time scale. There is, however, a younger and lively subfield of historical sociolinguistics, also called sociohistorical linguistics. It is essentially the attempt to conduct the same kinds of sociolinguistic analysis used for present societies on linguistic usage of bygone societies. It therefore relies on written records. It continually confronts the problem that written texts do not offer the same possibilities for investigation of social factors in language use as living observable communities do, a problem philologists deal with constantly and effectively.Footnote ³⁰

Contact linguistics, another young subfield, is concerned with what happens to languages when they are used in the same place at the same time. Contact linguistics is therefore also concerned with human relationships in a secondary way, but the primary goal of this research is to understand linguistic effects between languages simultaneously in use in the same situation, not human relationships or societies themselves. It is, in effect, an approach to language change between languages not accounted for by the methods of comparative historical linguistics. That said, the designation “contact linguistics,” referring to “language contact,” is technically sloppy, for a reason already mentioned: languages mostly “contact” each other in the brains and behaviors of multilingual individuals.Footnote ³¹ Research in this area continues to yield important new results, mostly focused on recent times, although a growing number of philologists have been taking the findings of contact linguistics into consideration in their accounts of ancient languages.

These three fruitful kinds of linguistics are primarily concerned not with history at large but with specific aspects of language and the nature of language itself. That is because they are parts of linguistics. They stand somewhat autonomously as subfields because of their respective successes. Although linguistic history is related to all these important concerns, its goal is history, relating the history of people with that of their particular languages and vice versa, not with language as a general phenomenon. Linguistic history does have the potential to inform the study of language as such, and it will do so with well-documented instances and examples. Likewise, it will benefit from the models and methods of general linguistics.Footnote ³² Linguistic history is like historical linguistics in its extensive diachrony, but unlike it in that it is rooted in the history of the speakers of a language, without which a language cannot exist or undergo change. In its focus on speakers and their relationships, linguistic history resembles sociolinguistics, but it is far more extensive in its diachrony. It necessitates the study of language contact, but its primary goal is not to discover universals of linguistic phenomena.

To summarize, linguistic history calls for explanatory pluralism but it stands on its own. It is like historical linguistics in studying language change over time, although it is different from it in addressing social change and the history of speakers simultaneously and inseparably from language change. It is also like sociolinguistics in that it attends to human relationships and society and their consequences for languages, but it is unlike sociolinguistics in using eclectic methods to access the speech and language use of speakers long gone and in its extensive diachrony. It participates in contact linguistics in studying bilingual and multilingual individuals and societies, as well as the linguistic effects due to such situations, but it is concerned with all aspects of linguistic diachrony, not with contact alone. Like all but the most recent history, linguistic history largely depends on written records of languages of the past before audio recordings of speech existed, commonly available only for about a century now. Therefore, philology in every sense of the term must be the primary methodological basis of linguistic history. This kind of linguistic history may even be considered a new branch of philology. Philology itself is not unitary; it is an umbrella term. Sometimes the name refers to textual criticism and the preparation of texts for publication and study, without which sound records for the study of language over long periods are unavailable. Sometimes it refers to comparative-historical linguistics itself, which is necessary to understand change in languages. Most broadly, it refers to the historical-contextual study of written texts, refusing to treat texts as abstracted objects lacking a contextual world that produced and preserved them. Linguistic history needs all other kinds of philology because these methods give the best access to premodern languages and speech before audio recordings. In this it is the same as historical linguistics. As Labov quipped, historical linguistics “can be thought of as the art of making the best use of bad data,” bad in the sense that these data are obtained not from living subjects whose living behavior is directly observable and through scientific methods of sampling, but through the vagaries of the preservation of the media of writing across the vicissitudes of time.Footnote ³³ Linguistic history is in the same situation. The combination of philological methods makes the best use of such data. The power of linguistic history is due to reliably tested correspondences, one of the bases of all science: it can diagnose social history through historical linguistic data while it can also diagnose linguistic change through historical social data. (This will be explained in Chapter 2.) It does so with the models of various subfields of linguistics, at least a dose of sociology, and the methods of history and philology.

Because it necessitates explanatory pluralism, it is not always necessary to demarcate linguistic history very sharply from these other kinds of linguistic research. It participates in them, and those kinds of linguistic research do sometimes contribute directly to history. The goal of linguistic history is to produce specific histories of languages in terms of the histories of their speakers, not to address universals and general tendencies – although it should contribute to the latter and it can benefit from them, too. The linguistic history offered here, as the reader will see, is of a specific kind. While enjoying the valuable proceeds of various kinds of linguistics, this book adds new methods and models for understanding language change in history as a function of changes in human society affecting their linguistic relationships. It relates language change over a long period to social, political, and economic change, and vice versa.

Last, linguistic history should avoid propagating myths about particular languages and about language in general. Linguistic history must historicize. Because language ideologies rely on tendentious historicization, linguistic historians must be conscious of these pitfalls and try to avoid them.Footnote ³⁴ Linguistic history should contribute rather to the demythologization of languages and the critical analysis of the stories told about languages invented by the people who speak them for their own ends.

Terminology

This project uses distinct terms deliberately in specific ways. The Persian language is one of a subgroup within the Indo-European language family commonly called Iranian in English. I refer to these languages rather by the term Iranic to distinguish the languages, spoken over millennia in many countries and regions, from the modern nation of Iran, its Iranian people, and Iranian culture. The term Iranic is not new and offers clear advantages for discussions of the languages involved.Footnote ³⁵ It aligns with the English terms used in adjacent fields: English is a Germanic language, not a German language; Uzbek is a Turkic language, not a Turkish language; Sinhala is an Indic language, not an Indian language. The different English adjectives help to keep the different kinds of phenomena clearly distinct, and they contribute to holding some ideological tendencies at bay.

When Christian Lassen, followed by other nineteenth-century linguists, identified the Iranic languages as a group with common features and designated them as such (iranisch in German), he was drawing on the ancient endonym of the speakers of the Iranic languages, Arya (Old Persian ariya), the keyword in the etymology of the name Iran.Footnote ³⁶ Lassen rightly distinguished the Iranic languages as a group specifically distinct from their cousin Indic languages of India, the speakers of which were also known as Arya. He called the Iranic languages iranisch without concern about confusion with the name of a modern state, because what we call Iran today was generally known as Persia in nineteenth-century Europe. Linguists of that time, however, assumed that the history of a language and its speakers was also a racial or national history. A language corresponded with a nation, one to one, it was assumed, and the speakers of a group of related languages constituted one race. In this connection, the idea of a distinct “Iranian people” who constituted or were a part of an “Aryan race” developed.Footnote ³⁷

Only in 1934–1935 did the government of King Reza Pahlavi request that foreign states refer to his country as Iran rather than Persia, a controversial decision at the time.Footnote ³⁸ The tendency thereafter was to connect the modern nation of Iran to a whole family of languages and to call all their ancient speakers “ancient Iranians” regardless of the land in which they lived, the specific language they spoke, or their different cultures. The “Iranian” languages were then more easily appropriated as an entire group for the nonlinguistic purposes of Iranian nationalism and for the modern idea of a distinct Iranian race of the Aryan “whites” of Asia. These conceptions left an imprint within Iranian studies still noticeable today, seldom addressed directly.Footnote ³⁹ As is well known, the term Aryan was furthermore appropriated for other political purposes and given almost irrevocable new moral connotations by the National Socialists of Germany and their Third Reich, as well as by some Iranian nationalists.Footnote ⁴⁰ Suffice it to say that the term “Aryan” now bears such inappropriate signification for the study of the ancient Persian Empire and other Persian-led polities of the past that its use should be discontinued for this purpose. When there is so much ideological baggage relevant to modern history but not to ancient history, it is necessary to parse the terms carefully. Ancient Persians like King Darius I called themselves Arya in their own language in a context having little to do with the modern term Aryan as it is used in English. By deliberately reverting to the ancient adjective Arya (without the English suffix -n), I employ the term with its ancient use. Therefore, I write occasionally about Arya languages and Arya people, which is, in effect, a shorthand way to say Iranic-language-speaking peoples – without referring to “Aryans” racialized in modern ways. In this context, it should be clear that I am not referring to the speakers of Indic languages who designated themselves Arya, either. The term Arya has the further small advantage of avoiding confusion with the Christian “Arian heresy” of the bishop Arius (d. 336), another homophone apt to trip up the unwary. Those more accustomed to discussing “ancient Iranians” and “Aryans” will easily understand my purposefully modified usage of terms.

The Old and Middle Iranic languages, all of which are first attested before the advent of Islam, are conventionally regarded as belonging to “ancient Iran.” The title of this book uses that convention to describe Old and Middle Persian together as Ancient Persian.

I use the words Achaemenids and Sasanids to refer to the members of their respective dynasties, the kings themselves and their descendants, treating the word ending -id as referring to the descendants of an individual. I use the words Achaemenian and Sasanian, with the more general adjectival suffix, to refer to things pertaining to their states, the people they ruled, and their cultures. For example, Xerxes was an Achaemenid, his empire Achaemenian.

I refer to this work as linguistic history. The adjective from that is linguistic-historical.

Plan of the Book

Chapter 1 introduces the linguistic phenomena exhibited by the Old Persian corpus of texts. It presents and summarizes what is well known to specialists but not widely known to ancient historians generally, and it introduces the major standing problem to be investigated here: how the ancient Persian language was drastically restructured within several generations. Chapter 2 lays out the linguistic-historical model that can explain the phenomena in a new way. In this long chapter I have synthesized that model, and the theory it represents, from various subfields of linguistics. The most important contributing subfield is contact linguistics, the branch of linguistics that addresses how the languages that people speak change when people acquire new languages under different circumstances and when they use more than one language in the same place. Modern cases of language restructuring are introduced to illustrate potential comparanda with Persian and for the use of other linguistic historians. Chapter 3 applies the linguistic model of Chapter 2 to the problem introduced in Chapter 1. It provides necessary historical contextualization to make sense of the data. This chapter takes a new look at long-known sources in Greek, Old Persian, Elamite, and other languages, and to a very limited extent also material culture preserved through archaeological research, to explain the linguistic history of Persian in this ancient period. It makes clear how the linguistic model presented in Chapter 2 applies to the Persian language in the time of the Achaemenian Empire and it addresses counterarguments. Chapter 4, the last, expands that application to other Iranic languages and to the historical context on a larger scale. It buttresses the findings of Chapter 3 by broadening the scope of the investigation, showing that Persian was not the only Iranic language affected by the social changes induced by the empire of the Achaemenid kings. The Conclusion summarizes the findings of this research and considers its ramifications.

The original plan of this book included a second, larger part concerning the genesis of New Persian and its evolution from Middle Persian. The social factors that conditioned the genesis of New Persian from Middle Persian are similar in some ways, but different in other, critical ways, to the factors that created Middle Persian from Old Persian. The origins of New Persian are also much better documented than those of Middle Persian, but the social history discernible from richer sources is correspondingly more complicated to analyze and to discuss. Together, the two parts of the original project grew too large to be accommodated in a single volume for today’s readers and the preferences of today’s presses. Although they both deal with the linguistic history of Persian using the same methods, they also deal with events separated by many centuries. They rely on different corpora of sources, written in different languages. For these reasons, and with the advice of various publishers who considered the project at its largest, I decided to separate them into two books. The second book, analyzing the evolution of New Persian, will appear in due course.