For centuries, language experts and casual learners alike have recognized Persian as a “simple” language, referring to the morphology that must be learned to speak or read it. As specialists in ancient Persian have known for some time, the Persian language took on most of this inflectional simplicity more than two thousand years ago. I have argued here that the transformation of Persian from a typically highly inflected ancient Indo-European language into one of the inflectionally simplest known ancient languages – the change from Old Persian to Middle Persian – was a side-effect of the Persian imperial society created by the Achaemenids. Their empire created conditions in which many adults who spoke many different languages needed to learn to speak at least some Persian with their new masters and even with each other in specific imperial sites. The unprecedented reach of the Persian Empire created conditions for intensive and extensive contact between native speakers of Persian and diverse foreigners who had to learn Persian language to participate in the new society either as elites or as their servants. The latter probably outnumbered the former in many places where Persian was current. Some of the most important sites of this intergenerational change in Persian inflectional morphology were the palaces and mansions of the Persians themselves. Persians populated their own homes, even their most private residences, with many foreign servants, slaves, eunuchs, and concubines taken from distant lands. The armed men who served the Persians likewise came from many different linguistic backgrounds and some of them needed to learn Persian to cooperate in that service. Persians probably simplified their own speech as foreigner talk, while a large wave of nonnative learners of Persian transformed the language by reducing its morphology to make it easier to speak as adults who did not grow up speaking Persian. This is the well-documented normal effect of such an event. It is likely that the influx into these sites of adults who needed to learn Persian never ceased for the duration of Achaemenian power. All later forms of Persian exhibit the same simplified morphology, demonstrating that the grammatical changes were not just the nonnative usages of foreigners, but that they were learned as native speech by later generations of Persians generally. Within decades, Persian children were learning Persian from nonnative speakers as much as from native speakers. In about five generations, the native varieties of Persian had lost most of the inflectional morphology that Persian had exhibited about the year 500 bce. By the time of Alexander’s invasion, it had become an early form of Middle Persian, or something close to the Middle Persian attested in later centuries.

The conclusion just stated is based on the application of a complex model of social factors in language change to a complex set of data. About the data there has been widespread agreement since the nineteenth century. It has long been noted that something unusual happened to some of the ancient Iranic languages, and particularly to Persian for which Achaemenian written records are available, as compared with other ancient Indo-European languages. The inscriptions of the Achaemenids provide the clearest testimony to this. Scholars have concurred that most of the Old Persian grammatical inflection had disappeared by the end of the Achaemenids’ dynasty, and that this indicated the genesis of Middle Persian. Although they have sometimes described this variously with evaluative terms current in the nineteenth and early twentieth centuries, such as “decay,” they were nevertheless correct that the change had substantially occurred already in the Achaemenian period and that it was therefore a rapid transformation. They also have sometimes agreed that Iranic languages other than Persian have been affected by a similar process before the times of their earliest attestation in writing, making some ancient Iranic languages seem to resemble “modern languages,” meaning, implicitly, those of contemporary Western Europe.

What has been missing is an explanation for the apparent grammatical aberrancy of Persian. Historical comparative linguistics, the main tool used in deciphering the ancient Iranic languages with remarkable success, has lacked models adequate to explain this event in the history of Persian and other Iranic languages. Instead, there have been only moral evaluations or tentative guesses, as well as a few intuitive hypotheses that language contact was somehow responsible. It was necessary here, therefore, to use complementary material synthesized from several sub-branches of linguistics to assemble a cogent model that could account for the rapid transformation of Old Persian into Middle Persian. These derived principles are much newer than the decipherment of Old Persian and are related to debates about general issues in linguistics. They have not been digested hitherto by most scholars of ancient languages, be they Iranic languages or others. Nevertheless, the principles are sound, and they are corroborated as complementary findings from related areas of linguistic research. The combination of the long-established inscriptional data and the labors of generations of philologists with the new linguistic-historical model synthesized here has created a historical account that is furthermore supported by what has long been known from familiar primary sources. The Achaemenian royal inscriptions in Persian represent perhaps the earliest clearly documented instance of semicreolization, a social process tied to patterns of abrupt population contact and large-scale adult language acquisition. In this case the process was induced by the use of imperial power to gather linguistically diverse subject peoples at sites presided over by Persian speakers, rapidly and naturally triggering a transformation of the language of the imperial masters through the nonnative varieties of Persian by the newcomers. It is often remarked that the Achaemenian Persian Empire was imperial on a scale far greater than that of any prior empire. “World empire,” a nonanalytical category, has often been invoked to convey the idea that the Persian dominion was unprecedented in scale. The coincidence between this common observation and the restructuring of the Persian language of the rulers is telling. It supports Meillet’s intuition, when discussing the Iranic languages, that “To become imperial is, for a language, the most serious of crises.”Footnote ¹ More precisely, “the most serious of crises” for the inflectional morphology of a language is a demographic upheaval in which mass adult acquisition of a language is necessary, although Meillet’s formulation sounds better.

Generations of historians of this Persian Empire have already intensively analyzed numerous sources about social arrangements in Achaemenian Persian society, so that the Achaemenian Empire is better known than many much later empires in human history. Here, the interpretation of some of these sources has been improved through the combined application of the specific linguistic data and the linguistic principles. Together they illustrate change in the constitution of the Persian ruling class, and they happen to shed some light on members of Achaemenian society that normally stand in the shadow of the kings who predominate in our narratives, almost completely ignored in the sources. This tells us something genuinely new. The assumption that only Persians spoke Persian and that there was no imperial melting-pot must be strongly qualified if not discarded. It is true that most of the subjects of the Persians had little close contact with the ethnic elite of the empire, as has been the case in most empires. Such people continued to work in their fields, tend their flocks, or do whatever it was that provided them with a living, and they also continued to speak the languages of their parents. But the view is mistaken if we look at the sites in which Persians themselves lived. We can infer confidently that workers, concubines, and eunuchs – newcomers, foreigners, people enslaved after being taken as tribute and as captive – played a large part in the reshaping of the language at the center of the elite Persian culture that they nurtured and the elite culture of this mighty empire. This research thus gives voice, albeit very indirectly, to those forced into servitude. The grammar of the language itself testifies to their presence in Persian mansions and palaces. It was the imperial power of the Persian kings, and the patriarchy and slavery that has prevailed in most complex societies, that subjugated these people and united them in royal sites. The Persians’ free enjoyment of that power over others inadvertently led their own language to evolve into a form that would be easier for adults to learn as they entered their service. The ensuing unilateral bilingualism – others had to learn Persian, but Persians seldom had to learn other languages – meant that speakers of other languages would adopt Persian vocabulary, but by contrast the Persian language would be grammatically reduced. Exposed to exoterogenic forces, Old Persian underwent semicreolization; it emerged as Middle Persian, an ancient Indo-European language with drastically reduced inflectional morphology. To say that Middle Persian is a semicreole of Old Persian is, however, only to classify it with respect to its origins. By the Sasanian period in the third century ce, after hundreds of years of use, Middle Persian’s origin in semicreolization was already ancient history. That is the character of languages transformed through episodes of morphological reduction induced by mass nonnative acquisition. They retain creole-like characteristics long after the formative population event has ceased. To use the term semicreolization here is to accept that creolization is a predictable phenomenon in language change consequent to uncommon but normal social changes, and moreover that languages do not belong to two sets, creole or noncreole, but that degrees of restructuring are possible. As explained in Chapter 2, to invent another term for this phenomenon would obfuscate the relationship between the effects of a demographic upheaval in the history of Persian and the effects of similar events on other languages. Persian is a language like any other, susceptible to the same processes as other languages in specified conditions.

In his book The Growth and Maintenance of Linguistic Complexity, Östen Dahl posited that linguistic complexity, such as ornate inflectional morphology, “may even be negatively correlated with the rise of large-scale societies with highly mobile populations.”Footnote ² This book supports that informed hypothesis with a major case study from ancient history. When the ancient Persians created an empire necessitating long-distance mobility and sudden population mixing on a scale unprecedented in human history, they triggered drastic effects in their own language. Persian became, in a word, imperial. This event also explains why some specialists in Iranic languages – as discussed in Chapters 2 and 4 – have been inclined to see Middle Persian as virtually a “modern” language. Their intuition correlates with Dahl’s hypothesis in that the “modern” languages that one has in mind for such comparisons exist in the wake of more recent extensive empires that have facilitated mobility on a truly global scale, involving a larger population of humans than has ever been sustained on earth before. This suggests that one may speak loosely of “imperial languages” under certain circumstances. Such a term is not an analytical category. It is descriptive of a tendency at best, because not all languages of imperial elites will undergo the same effects. In the case of the ancient Persian language, nevertheless, it seems quite appropriate.

Trudgill reflects that “in the future, we are increasingly unlikely ever again to see the development of highly inflectional, fusional language varieties” with very complex morphology – at least in the foreseeable future.Footnote ³ That is because we live in a period of increasingly global communication. We benefit and suffer from networks of exchange that extensively surpass any sort of imperial pattern hitherto known. We are witnesses to mass migration and population contact in many directions on a scale never seen before. Beyond the large-scale relocations of people themselves, electronically mediated nonlocal or long-distance relationships are replacing many of the face-to-face interactions hitherto normal for all people of history. The intimate societies of isolated groups and thick, dense social networks based on physical propinquity, which ensured constant small-scale conflict between little self-supporting communities, are eroding in different places at different rates. They are being exchanged for what seem like surrogates for the relationships in which humans have evolved, along with inequalities vaster than any known before and conflicts now on global scales. Following the lines of linguistic history, we could point at the empire of the Achaemenians as one of the earliest monumental landmarks on this path toward mass migration, population scrambling, and pervasive open social networks, with their expected linguistic effects. The case of the Old Persian language shows us that the first so-called world empire of the Persians was a harbinger of these social changes that predictably leave a language with reduced morphology in its wake, shaped by the forces of functionality to satisfy the urgent need of adults to learn it. It is with the Achaemenian Persian Empire that the correlation of patterns of grammatical reduction with a demographic upheaval first comes into clear historical view. It will be interesting if older examples of this syndrome can be detected by those studying still earlier language records, but it is fair to say that it is the mighty Persian Empire that marks for us the beginning of a new age of imperial demographics tumultuous enough to leave traces even in the grammar of its rulers’ language.Footnote ⁴

More precisely, I have made the case that ancient Persian exhibits the oldest clearly documented instance of the semicreolization of a language, the drastic but incomplete reduction of its inflectional morphology, due particularly to the imperial status of its users and the demographic mixture induced by extensive and intensive coercion – with all the caveats and pointers supplied about the term semicreole provided in Chapter 2. The evolution of Persian from the fourth century bce until today has retained the clear signs of that event ever after in the relative learnability of its morphology, a vestige of ancient Persian imperial culture still reaching students in Persian classrooms today.