the sources. As recently as the late twentieth century, it was accepted historical knowledge that the fall of the Assyrian Empire was followed by the rise of a Median “empire” which ruled vast tracts of the Ancient Near East for half a century, until Astyages, the last Median ruler, was overthrown by one of his own vassals, namely Cyrus the Great. Only relatively late, the important works of the late Heleen Sancisi‐Weerdenburg pointed out the many difficulties and inadequacies of this view (Sancisi‐Weerdenburg 1988, 1995; see also Kienast 1999). Sancisi‐Weerdenburg was particularly critical of the alleged “imperial” structure and character of Median rule and identified a number of striking dissimilarities with other imperial entities of the Ancient Near East. She also emphasized the almost complete dependency of modern historiography on Classical (i.e. Greek) sources, to the nearly complete disregard of Ancient Near Eastern sources. Unfortunately, Sancisi‐Weerdenburg's work was met with very little acceptance. On the contrary, her hypotheses and conclusions were ignored, and the problematic nature of previous scholarship was marginalized.
An international symposium held in Padua in 2001 attempted to rigorously review all available sources and to present a secure (as far as possible) narrative of Median history (Lanfranchi et al. 2003). While the participants were largely successful in their first aim, no consensus on an accepted narrative could be reached due to the frustratingly incomplete and fragmentary nature of the sources. There was, however, a general consensus that the existence of a Median “empire” cannot be conclusively proven and should not be treated differently from other hypotheses. Opinions were divided on any further detailed characterization of historical events: whereas many of those present rigorously refused it or attempted to follow through on the theses of Sancisi‐Weerdenburg by questioning the geographic extent of Median influence (Liverani 2003; Rollinger 2003a, b; Henkelman 2003; Jursa 2003), others still chose to accept the notion of a Median “empire” (Roaf 2003). The discussion has continued up to the present (Tuplin 2005; Lanfranchi 2021; Rollinger 2004, 2005, 2009, 2010, 2011, 2020a; Waters 2005). What follows is an attempt to tackle this “Median dilemma”1 and to throw a more general light on the Medes and their “history,” the available sources, the problems, and what we can state with some certainty.
In the reign of the Neo‐Assyrian king Shalmaneser III (858–834 BCE), Medes (Madāya) are mentioned for the first time. In the following centuries, they appear again and again in Neo‐Assyrian sources, especially in royal inscriptions, but also in archival records, mainly as opponents encountered when the Assyrian armies campaign in the central Zagros area – where they are primarily localized – but also as vassals of the Neo‐Assyrian super power (Radner 2003; Bagg 2020, pp. 379–382). Although it is unclear how far to the east the Assyrians reckoned with the presence of a Median population, there is evidence that it was as far as the region of the modern cities of Teheran and Rey (Rollinger 2007).
Not only for the Neo‐Assyrian era of the ninth through seventh centuries BCE but also for the following Neo‐Babylonian and early Persian times (sixth century BCE) our sources exclusively exhibit an external view of the Medes (Liverani 2003). There is not a single indigenous source representing a “Median” perspective on their matters, their history, or their agenda. Nor do we know whether there was a shared Median identity and whether the Medes of our sources called themselves Medes (Lanfranchi 2003: p. 84). Sometimes the Neo‐Assyrian sources refer to “mighty Medes” and “distant Medes.” At least these qualifications look very much like projections from outside in order to organize the expanding knowledge of an area becoming increasingly well known by the Assyrians. Some of these “Medes” were localized inside the empire, some of them outside; this makes them, as seen through an Assyrian lens, a border population. The Medes within Assyrian reach were regarded as vassals and had to swear the loyalty oath to the Assyrian heir apparent Esarhaddon (672 BCE).
The origin of the term Madāya is unknown, its specific trans‐regional usage evidently derives from Assyrian practice. Like in antiquity the ethnic term “German,” picked up from a very local and indigenous usage and artificially spread over the entire population east of the Rhine river by Caesar himself, the Assyrians might have taken up a local designation somewhere in the central Zagros area and transferred it to a far larger population covering the whole of the central Zagros and farther to the east.
Although the most popular one, the designation Madāya/Medes was not the only one used for the population of the central Zagros area. In the second half of the eighth century BCE appears the designation “Arabs of the east” (Radner 2003: p. 55). We do not exactly know what this actually means but it seems to reveal some kind of uncertainty about how to label the peoples of the central Zagros regions. “Arab,” a term that also appears for the first time in world history in the inscriptions of Shalmaneser III, may, at least initially, represent not an ethnically or linguistically determined designation but one referring to a specific mode of living, where transhumance or trade with camels might have played a major role (Lanfranchi 2003).
In Neo‐Babylonian and (retrospective) Persian sources the term Ummān‐manda appears for the Medes (Adali 2011). The term clearly is a designation deriving from outside and has a pejorative connotation. Moreover, it is evident that the term Madāya and the Neo‐Assyrian concept attached to it – i.e. a rather homogeneous and substantial population of the central Zagros area – became part of a tradition and was adapted by contemporary and later, adjacent and more distant languages and cultures, like the Urartians, Babylonians, Hebrews, Greeks, and Romans.
As we do not know whether these Madāya had a supra‐regional Median identity, we are also ignorant about whether they represent a homogenous linguistic group. Although the many “Median” proper names that we have, thanks to the flourishing cuneiform and later especially Greek sources, reveal a dominant Iranian background of these people, one should be cautious about claiming that this evidence definitely proves a homogenous and well‐defined Iranian language, generally and simplistically labeled “Median.” Such a hypothesis, although very common and likewise broadly accepted as fact, does not rest on firm ground (Schmitt 2003; Rossi 2010, 2017). It is highly probable that behind these proper names lurks a much larger and more complex diversity of local Iranian dialects/languages. And certainly, we have to reckon with a larger ethnic diversity in general in these areas where Urartian, Hurrian, Elamite, Assyrian, Babylonian, and other languages played a certain role (Fuchs 2011).
Over the 200 years of rather extensive Neo‐Assyrian documentation there is not a single piece of evidence for a unified political entity in the central Zagros region, let alone for a Median “empire.” Instead, these sources depict a highly fragmented political landscape in the Zagros mountain region, with no discernible tendencies toward greater centralization. The Assyrians encounter a plurality of small political units, whose rulers they call not “king” but “city lord” (bēl āli). Although transhumance was highly important in these regions, horse breeding and trade/robbery also played an important role, since the Khorasan road, i.e. the predecessor of the Silk Road, crossed the central Zagros area. In any case, the Assyrian sources never describe the Medes as nomads but always as a sedentary population. Intensive contact with the Assyrian super power and its gigantic economic space was highly influential in these regions, especially when with the reign of king Sargon (721–705 BCE) the Assyrians began to establish provinces. These contacts certainly transformed local societies and may also have triggered developments toward political unification that can be described as “secondary state formation” (Brown 1986; Rollinger 2020b), although this process never became a supra‐regional phenomenon. This conclusion is confirmed by archeological sources that, likewise, do not indicate the existence of a unified Median state. Important sites like Nush‐e Jan, Godin Tepe, Baba Jan Tepe, or Ozbaki Tepe do not represent “imperial centers” but rather seats of “city lords” with no more than a local reach (Liverani 2003; Stronach 2003; Gopnik 2011). Previously identified seats of power of an alleged Median “empire” in western Iran, like Hamadan, or outside the proper central Zagros area, like Kerkenes Dag˘ı in Asia Minor, do not hold up to critical scrutiny and have been revealed as optimistic academic mirages, constructed to fit preconceived notions of imperial Media (Boucharlat 1998; Sarraf 2003; Rollinger 2003b).
By about the middle of the seventh century