‘Spirits of the Dead’ or ‘Necromancers’? The eṭemmū in an Old Assyrian Letter Reinterpreted in Light of Hebrew ’ōbôt, yidd^e‘ōnîm, and ’iṭṭîm

Abstract

The Old Assyrian archive from Kanesh, dated to ca. 1950–1850 BCE, has yielded a letter that refers to the consultation of the spirits of the dead (eṭemmū), thus making it the world’s oldest actual attestation of necromancy. However, whereas the immediate context mentions the šā’ilātum, ‘the women dream interpreters’, and the bāriātum, ‘the women omen interpreters’, a necromantic professional is lacking in relation to the questioning of the eṭemmū. Earlier studies have explained this discrepancy by suggesting that necromancy was part of the skill set of the aforementioned female professionals, or that the communication with the spirits happened directly, without the immediate involvement of a skilled specialist. The present article rather argues that the term eṭemmu, ‘spirit of the dead’, had a wider semantic range than hitherto held. In rare cases, it could also designate a necromancer. This proposal is supported by an identical semantic phenomenon in another ancient Semitic language. The biblical Hebrew terms ʼōbôt and yidd^e‘ōnîm not only refer to the spirits of the dead but also to necromancers. The same might be argued for the apparent Hebrew cognate of Akkadian eṭemmū, the hapax legomenon ’iṭṭîm in Isaiah 19:3. On the strength of the findings presented in this study, it is concluded that the fleeting blending of the spirit with the necromancer lies at the heart of this semantic merger.

1. Introduction

Despite the abundance of magico-divinatory texts in Mesopotamian written culture, there is comparatively little evidence of necromancy. However, one of the earliest references to the consultation of the dead is attested in archival materials from Kanesh (Kültepe), which date from between ca. 1950 and 1850 BCE. This private archive contains an Old Assyrian letter (TC 1,5 = TCL 4,5) addressed to the merchant Imdī-ilum, in which he is urged by two female relatives to return home because the god Aššur has warned that his life is in danger. Aššur relayed this message via the šā’ilātum, ‘the women dream interpreters’; the bāriātum, ‘the women omen interpreters’; and the eṭemmū, ‘spirits of the dead’, whom Imdī-ilum’s sisters had consulted. Although the bilingual Lu₂ lists offer several terms for necromancers in the Akkadian language, none of them are employed in this letter. The lack of reference to a necromantic specialist has raised the question in previous scholarly studies whether the women themselves consulted the spirits of the dead, an interaction fraught with danger, or gained access to them via the šā’ilātum or the bāriātum. This study explores a tantalizing alternative possibility, namely that the term eṭemmu1 had a wider semantic range than previously held and may in rare cases have denoted a necromantic professional. A semantic similarity is attested in the Hebrew Bible, wherein the terms ʼōbôt and yidd^eʻōnîm not only refer to the spirits of the dead but also to necromancers. Finally, this study examines whether the same ambiguity applies to the apparent Hebrew cognate of Akkadian eṭemmū, the hapax legomenon ’iṭṭîm in Isa. 19:3.

2. Imdī-Ilum Receives an Alarming Message from Aššur

In the 1920s, the Czech orientalist Bedřich Hrozný made a remarkable discovery: the correspondence of the businessman Imdī-ilum, who had settled in the Assyrian trade colony Kanesh (modern-day Kültepe). These archival materials, dating from the first half of the nineteenth century BCE, bear witness to the dynamic professional life and family ties of an otherwise marginal character in Old Assyrian history (Larsen 1982). The unearthed and deciphered clay tablets mostly deal with Imdī-ilum’s business transactions, from which we glean that he was a wealthy, successful merchant who traded in tin and textiles. On a more personal level, we learn rather little about Imdī-ilum. However, an intriguing letter sent by two women gives a revealing insight into one of his character traits. Tarām-Kūbi and Šīmat-Aššur, who may have been Imdī-ilum’s sisters (Michel 2020, pp. 426–34), warn him that his desire to accumulate wealth has put his life in danger and urge him to return home, to the capital city Aššur. The respective tablet reads as follows:2

Obv. ¹a-na Im-dí-dingir qí-bi-ma ²um-ma Ta-ra-am-Ku-bi₄-ma ³ú Ší-ma-at-A-šùr-ma ⁴a-na-kam ša-i-lá-tim ⁵ba-ri-a-tim ù e-ṭé-me ⁶nu-ša-al-ma ⁷A-šùr uš-ta-na-ad-/kà lo.e. ⁸kù-babbar ta-ra-am ⁹na-pá-áš-ta-kà rev. ¹⁰ta^!(ZA)-ze-ar i-na ¹¹a-lim^ki A-šùr ma-ga-ra-/am ¹²ú-lá ta-le-e ¹³a-pu-tum ki-ma ṭup-pá-am ¹⁴ta-áš-me-ú al-kam-ma ¹⁵e-en A-šùr: a-mu-ur-ma ¹⁶na-pá-áš-ta-kà u.e. ¹⁷e-ṭé-er ší-im túg^tí-a le.e. ¹⁸mì-šu-um ¹⁹lá tù-šé-ba-lam

^1–3 Say to Imdī-ilum: thus (speak) Tarām-Kūbi and Šīmat-Aššur.

^4–6 Here (in Aššur) we consulted the women dream interpreters, the women diviners, and the spirits of the dead, and (their answer was: ⁷ the god) Aššur keeps on warning you; ^8–10 you love money (so much that) you despise your own life! ^10–12 Can’t you comply with (the god) Aššur’s (wishes here) in the city (of Aššur)? ^13–14 Urgent! When you have heard the letter, (then) come here, ¹⁵ meet (the god) Aššur face to face, and ^16–17 save your life!

^17–19 Why don’t you send to me the proceeds from my textiles?

Figure 1 below shows the clay tablet.

Although the letter is dispatched by Tarām-Kūbi and Šīmat-Aššur, it appears that the former is the main correspondent, judging from the shift in voice to the single person in lines 17–19. Šīmat-Aššur may have been the youngest of the two sisters as she is mentioned as the second sender. The emotional tone of the letter is striking: the sisters are deeply concerned about Imdī-ilum, whose life is in danger because of his avarice. Their concern about his tireless pursuit of riches and his forsaking of his religious duties has even prompted them to consult various female diviners and the spirits of the dead, all of whom relayed the same message from the god Aššur. Imdī-ilum is in such a precarious situation that the sisters urge him to return to the city Aššur immediately and heed his religious duties. The supernatural flows seamlessly into the mundane when Tarām-Kūbi concludes the letter by querying her brother about the lack of proceeds from her textiles.

Garelli (1979) sought to reconstruct the events that led to the desperate plea of Tarām-Kūbi and Šīmat-Aššur. According to him, their letter fits chronologically in a chain of correspondence that involved Pūšu-kēn, another renowned businessman from Kanesh. Pūšu-kēn’s sister Tarīš-mātum and Bēlatum, who was presumably her daughter, urge him in increasingly desperate letters to meet his financial obligations, allegedly towards Imdī-ilum, because the god has struck Bēlatum with a severe illness, and they are tormented by demons (utukkū) and spirits (eṭemmū). According to a subsequent letter, illness has also befallen the wives of two of Pūšu-kēn’s employees, one of whom happens to be Puzur-Ištar, the son of Imdī-ilum. The women further write that they consulted the šā’ilātum to learn the reason for the misfortune that was striking Pūšu-kēn’s extended household. Garelli interpreted the repeated figurative use of the term abūni ‘our father’ by Tarīš-mātum and Bēlatum not as an honorific designation of Pūšu-kēn, but rather of Imdī-ilum, whose assets the women seek to recover, though none of their letters explicitly mention him. Any further delay in the settlement of this financial matter will prolong the divine rage. Hence, Garelli concludes, Tarām-Kūbi and Šīmat-Aššur decide to intervene and write to Imdī-ilum directly, resulting in the letter that is the topic of the present study.

In my view, the thematic similarity notwithstanding, there appears to be no direct link between the events described by Tarīš-mātum and Bēlatum on the one hand and by Tarām-Kūbi and Šīmat-Aššur on the other. A close reading of the letters involving Pūšu-kēn rather brings his religious misdemeanor to the fore (Michel 2020, pp. 373–76, esp. nos. 254–55). Moreover, he clearly is the sole recipient of the honorific designation abūni in these letters. Both businessmen are accused by female relatives of cultic negligence and financial misbehavior. In the case of Pūšu-kēn, we are even aware of the dire consequences that followed from his wrongdoing. Nevertheless, such pleading letters were not unique, as demonstrated by Michel (2020, pp. 359–60, nos. 236–37). The excavations in Kanesh have yielded various other letters in which women express their concern about the forsaking of cultic duties by their next of kin.3 Despite the apparent ties between both merchants’ families (cf. Michel 2020, pp. 431–32, no. 288), the divinely orchestrated woes that befell Pūšu-kēn’s immediate circle seem unrelated to the crisis that Imdī-ilum’s sisters referred to in their letter.

The intertwining of daily matters and the extramundane is not unique in the unearthed correspondence from Kanesh. Many of the female correspondents, who were mostly residing in the mother city Aššur, had to keep their households running during the oft-prolonged absence of their husbands. They had to provide for themselves and their children, pay off creditors, act as lenders, keep a watchful eye on their spouses’ business documents, and produce the textiles that were sent to Kanesh. On top of these considerable responsibilities, the merchants’ women had to fulfil their religious duties and appease the gods and the ancestors, some of the latter being buried in tombs underneath their family homes (Michel 2020, pp. 356–57, 400 n. 19). Hence, the dead were literally and figuratively ever present in the lives of these women.

3. The Quest for a Female Necromancer

Although other tablets from the private archives in Kanesh also refer to the eṭemmū, as we already witnessed in the correspondence addressed to Pūšu-kēn, the letter sent by Tarām-Kūbi and Šīmat-Aššur is unique in that it contains the first reference to necromantic praxis in Mesopotamian society. The sisters not only consulted the šā’ilātum and bāriātum, female diviners who would interpret dreams or omens, respectively, but they also resorted to the eṭemmū. It is interesting that although two types of female diviners are mentioned explicitly by their respective titles, the sisters do not refer to the title of a necromancer. Following on from the preceding female divinatory professionals, we would expect the professional designation of a woman necromancer. The bilingual Mesopotamian lexical lists of professions contain one titular reference to a female necromancer, both in Sumerian and Akkadian. According to the canonical Lu₂ list,

SAL IGI.ŠID-e₁₁-e-dè   ₌ mu-še-[li]-tum ‘(female) necromancer’, lit. ‘one who raises (the spirits of the dead)’4

The sisters make no mention of a mušēlītum in their letter, nor has this title hitherto been attested in another Mesopotamian text. The same holds true for the various designations for male necromancers, whose titles are limited to lexical lists: ša eṭemmi, mušēlû eṭemmi, and, possibly, mušēlû ṣilli (Tropper 1989, pp. 58–62).5

The lacking reference to a necromantic professional could indicate that the women directly consulted the eṭemmū (cf. Hirsch 1972, p. 72).6 Before we explore this possibility further, it might be useful to examine the meaning of the term eṭemmu, thereby following the studies by Scurlock (2016) and Steinert (2012, pp. 295–384). The eṭemmu is comparable to our modern-day idea of a ‘ghost’, although there is a stronger corporal aspect to the concept of the eṭemmu. Therefore, it might be better to conceive of the eṭemmu as a ‘body spirit’, lingering perpetually in the bones, which were considered the essence of the mortal’s body. Because of its presumed corporality, the eṭemmu could suffer from hunger and thirst in the netherworld, and therefore, the relatives had to continuously provide funerary offerings, thus ensuring that the deceased would remain safely confined to its resting place. In return, the eṭemmu could protect the living, act as their moral guardian, and intercede on their behalf in the supernatural realm. However, if the funerary offerings were interrupted or a person had suffered an untimely or violent death, for instance, through an accident, murder, or illness, the eṭemmu could launch an attack on the living, thereby wreaking great havoc. The misfortune that struck a family could also result from the activity of the utukkū-demons. When the memory of the deceased had long faded, the dead lost their individual nature and turned into these demonic entities, which are sometimes mentioned in the same breath as the eṭemmū. Both supernatural beings were capable of haunting the houses of the living and causing all kinds of illnesses, even death.

These various facets of the eṭemmū are traceable in the Old Assyrian archive from Kanesh.7 For instance, a woman urges her brother to send her the proceeds from the textiles. By averting her impending financial ruin, the brother will please the gods and the spirits of the dead (Michel 2020, p. 359, no. 236). In another letter, a woman writes that a long overdue debtor has upset not only the family but also the eṭemmū (ibid., pp. 420–21, no. 280). That the eṭemmū, alike the gods, are keenly aware of the (im)moral behavior of the living becomes also clear when a woman writes that the spirits of the dead can vouch for her innocence (ibid., pp. 455–56, no. 310). We already encountered Tarīš-mātum and Bēlatum, according to whom a god (Aššur?), the eṭemmū, and the utukkū-demons are plaguing Pūšu-kēn’s household because of his cultic negligence. In their divine and demonic wrath, they have even bestowed illness upon several female members of his extended family (ibid., pp. 373–75, nos. 254–55).

From the Kanesh archive, we learn that the women in Old Assyrian society sought to appease and honor the ancestral spirits by making continuous offerings, heeding their cultic duties, and living a morally steadfast life. They seemed acutely aware of the capricious and hazardous nature of the eṭemmū. The spirits had to be carefully and respectfully dealt with; any direct interaction of the living with the eṭemmū was fraught with danger. Hence, it seems unlikely that the sisters Tarām-Kūbi and Šīmat-Aššur consulted the eṭemmū directly to learn Imdī-ilum’s fate, without the involvement of a skilled intermediary. Alas, due to a lack of source material, we are left in the dark as to the actual manner of necromantic praxis in the Old Assyrian period. According to the scarce evidence from the first millennium BCE, necromancy seems to have been a largely institutionalized affair in Mesopotamia, practiced by ritual specialists, possibly even within royal circles.8

The legitimacy of Ashurbanipal’s claim to the Assyrian throne may have been decisively settled by a spirit. Ashurbanipal was the preferred successor of King Esarhaddon (r. 681–669 BCE), but there were other legitimate claimants to the throne. To further strengthen his candidacy, Ashurbanipal seems to have resorted to necromancy so as to seek the favor of the deceased Queen Mother, Naqi’a.9 This royal spirit confirmed that the gods Aššur and Shamash had ordained Ashurbanipal to be Crown Prince, and she bestowed her blessing upon him, thereby expressing the wish that his descendants were to rule over Assyria. However, our knowledge of this necromantic consultation is not based on an eyewitness account. A letter that may have circulated in royal circles preserves Ashurbanipal’s purportedly verbatim quotation of the Queen Mother’s message. Nevertheless, the tablet (K. 1152) is badly damaged, resulting in various interpretations of the letter, not all of them being overtly necromantic (Finkel 1983–1984, pp. 2–3; 2021, pp. 224–27; Tropper 1989, pp. 76–83).

Whichever way Ashurbanipal managed to claim the throne, the fact is that the library that he founded after his royal ascension held a Neo-Babylonian necromantic instruction manual (K. 2779). From the sixth and fourth centuries BCE, respectively, we have two Late Babylonian manuals (BM 36703, W. 22758/2). These three tablets, all of southern Mesopotamian provenance, detail the rituals and spells that were required to conjure up, control, and communicate with an eṭemmu (Finkel 1983–1984, 2021, pp. 228–43; Scurlock 1988, pp. 103–12, 318–42; Tropper 1989, pp. 83–103). According to one of the instructions in BM 36703, an intricate ointment—a mixture of oil, crushed animal remains, dust, and an upturned potsherd from a crossroads—ought to be left overnight and subsequently applied unto a human skull to summon the spirit, which was brought up with the help of the sun god Shamash. Once anointed, the skull, in which the eṭemmu temporarily dwelt, was ready for questioning by the necromancer, whose face was smeared with the same substance.10 It remains unclear whether the skull once belonged to the interrogated spirit or could have been any random human skull. In the absence of a skull, a figurine representing the spirit could serve as a substitute in this necromantic ritual.

These three necromantic manuals date from halfway through the first millennium BCE and originate from southern Mesopotamia. They are far removed, both chronologically and geographically, from our Old Assyrian letter. We can therefore only speculate on whether the type of necromancy practiced in Aššur in the nineteenth century BCE would have been remotely similar to the ritual ceremonies outlined in these tablets. Tropper (1989, pp. 102–3) argues that the performance of elaborate necromantic rituals would have been restricted to the upper echelons of Mesopotamian society. The expertise of the ritual specialists was sought after by politically influential people, royals even, as in the case of Ashurbanipal. Moreover, according to Tropper, it is no coincidence that these necromantic manuals are of a relatively late date in Mesopotamian history. Their attestation coincides with the written fixation, the literalization, of magico-divinatory incantations and rituals that surged from the Neo-Assyrian period onwards. It is unclear to what extent these necromantic spells and rituals derive from older practices and the social setting thereof. However, throughout ancient Mesopotamian history, there seems to have been a genuine apprehension of the risks involved in dealing with the eṭemmū. It is therefore probable that Tarām-Kūbi and Šīmat-Aššur sought professional assistance. If necromancy was a costly affair, as it most likely was in the later Neo-Assyrian period, the women would have been able to afford it because they belonged to the commercial elite in Aššur and were financially well-off.11

Given the lack of reference to a necromantic specialist in their letter, could the šā’ilātum and bāriātum have had any role to play in the sisters’ search for answers from the eṭemmū? In his discussion of our letter, Veenhof (1983, pp. 88–89) identifies the šā’ilātum as female dream interpreters, although he does not rule out their ability to enquire of the gods through extasy or incubation. Likewise, he continues, the bāriātum may have practiced not only hepatoscopy, but also other kinds of omen divination, such as those involving heavenly manifestations, oil, smoke, or flour. Veenhof does not refer to necromancy as part of the women’s skill set.

Oppenheim (1956, pp. 221–23) argues that the šā’ilātum not only interpreted dreams but also summoned the spirits of the dead, exclusively on behalf of fellow women.12 As supporting evidence, he cites our letter and another Old Assyrian letter,13 according to which ‘the šā‘iltu-priestess’ (sic) conveyed the message from an ilum. Oppenheim interprets ilum in this letter as ‘spirit’, rather than ‘god’.14 In addition, in the Lu₂ lists, the šā‘iltum is listed in proximity to necromantic terminology, which strengthens his view that she was involved in the consultation of the dead. Oppenheim considers her necromancy a marginal activity, not formally institutionalized in the Old Assyrian period. The frequent designation of this female diviner in the plural form, šā’ilātum, could indicate that they lived outside the official temple domain as a collective, clandestinely engaging in necromancy. He even asserts that their involvement in necromantic praxis lowered the social status of the šā‘il(t)um. Referring to the letter sent by Tarām-Kūbi and Šīmat-Aššur, Koch (2023, p. 174) states that it ‘clearly suggests that at least one, and perhaps both, of the diviners consulted the spirits of the dead’. Another angle on the discussion is offered by Finkel (2021, p. 222), according to whom these women practitioners might have been preferred if a client wanted to consult a female spirit. The argumentation brought forward by Oppenheim, and subsequently by Koch and Finkel, is alluring but remains debatable. The šā’ilātum and the bāriātum are regularly referred to in the Kanesh archive, but none of the other letters attest to their being involved in necromancy (Michel 2020, pp. 372–73, no. 253; pp. 336–37, no. 226; p. 375, no. 256); even our letter is very much open to interpretation. Moreover, the same line of thinking is not applied to the juxtaposition of the bāriātum and the utukkū-demons in a letter sent by Bēlātum, the aforementioned, presumable niece of Pūšu-kēn, to Šalimma (Michel 2020, pp. 372–73, no. 253). Although this tablet is only partly legible, Bēlātum seems to berate Šalimma for pressuring her into a consultation with the bāriātum and the utukkū in the city of Aššur. It is questionable whether the bārûm, ‘the male omen interpreter’, and his female counterpart, the bārītum, would have dealt with the utukkū as part of their divinatory skill set. An exorcist would rather have presided at a ceremony involving the capricious, dangerous utukkū. Hence, in her letter, Bēlātum most likely refers to two separate ritual specialists, one dealing with omen interpretations and the other with exorcism. Bearing Bēlātum’s message in mind, we should exercise caution when interpreting the juxtaposition of the šā’ilātum, the bāriātum, and the eṭemmū in our letter. Tarām-Kūbi and Šīmat-Aššur most likely resorted to three separate kinds of divination, practiced by three different ritual specialists. The šā’ilātum and the bāriātum disappear from the literary corpus after the Old Assyrian and Old Babylonian periods, and hence, we are left without any further traces of them and their praxis.

Alternatively, Tarām-Kūbi and Šīmat-Aššur may have wanted to veil the fact that a male professional had visited the family home. In his discussion of Ashurbanipal’s consultation with the deceased Queen Mother, Finkel argues that ‘the necromantic procedure could only have taken place in the building that housed the queen’s tomb, most probably within the tomb chamber itself’ (Finkel 2021, p. 227). If necromancy in the Old Assyrian period had the same requirement, namely, that the ceremony had to be conducted at the spirit’s burial location, the necromancer may have had to visit Imdī-ilum’s family home. In lieu of a female professional, a male specialist may have presided at the ceremony, presumably under a chaperone. Perhaps the sisters wanted to disguise this visit of a male stranger in their letter and neutrally referred to the etemmū instead.15 However, from what we have hitherto gleaned from Old Assyrian society, many of these merchants’ wives and female relatives were necessarily fiercely independent. Due to the prolonged absence of their husbands, fathers, and brothers, they had to conduct business on their behalf with non-related men. The Kanesh archive attests to visits from non-related men to women’s homes for business purposes (for example, see Michel 2020, pp. 326–27, no. 218). Therefore, the fact that a male necromancer had visited the family home might not have alarmed Imdī-ilum upon receipt of his sisters’ letter.16

Finally, it is worth exploring Tropper’s observation that the client of a necromancer might have directly communicated with an eṭemmu during a ritual ceremony (Tropper 1989, p. 61). He notes that despite the various professional designations for a necromancer in the lexical Lu₂ lists, this type of diviner is never referred to as šā’il eṭemmi, ‘enquirer of the dead’. The necromancer might have only been in charge of the ritualistic aspect of the ceremony, whilst the client themself would engage in the actual consultation of the eṭemmu. Tropper therefore interprets the Mesopotamian necromantic professional as being a ‘Totenbeschwörer’, i.e., a summoner of the dead, rather than a necromancer because the latter term gives the impression that the specialist’s main task was to communicate with the dead. The ‘Totenbeschwörer’ rather focused on magically and ritually conjuring up and controlling the spirits of the dead and protecting vulnerable people, especially ill ones, from their hazardous and devious powers. Building on Tropper’s observation, we could argue that Imdī-ilum’s sisters indeed consulted a necromantic ritual specialist but referred in their letter to their own direct interaction with the eṭemmū during the ceremony, which was presided over by the necromancer. The aforementioned letter from Bēlātum to Šalimma might offer support for this argumentation as it seems unlikely that Bēlātum would have sought direct communication with the utukkū without the presence of an exorcist.

4. A Semantic Shift: The Eṭemmu as Necromancer

Up to this point in our discussion, the most probable scenario is that the sisters inquired of the dead themselves, possibly at the deceased’s actual burial location, under the watchful eye of a necromantic ritual specialist. We should allow for another possibility, namely that the term eṭemmu had a wider semantic range than previously held and could have been a designation for a necromancer.17

As discussed earlier in the present study, the various designations for necromancers that feature in the Lu₂ lists seem to have been restricted to lexical use only. If Imdī-ilum’s sisters had been aware of one of these designations, they would have likely listed it, in its plural form, alongside the šā’ilātum and the bāriātum in their letter. Instead, the sisters refer to the eṭemmū, which may have been a colloquial way of describing necromantic specialists. The women’s letters from the Kanesh archive are written in a vernacular type of language and contain grammatical and syntactical errors, all of which make it probable that Tarām-Kūbi and Šīmat-Aššur had also mastered the skill of writing, even at a basic level (cf. Michel 2020, pp. 428–29). This was not exceptional in Mesopotamian society; female literacy was attested beyond scribal and elite circles (Halton and Svärd 2017, pp. 35–36; Michel 2020, pp. 333–35).

The colloquial use of the term eṭemmu as a reference to a necromancer could have developed in an era when necromancy was not yet associated with the intricate rituals and incantations that we know from the Neo-Assyrian period onwards, a time when the necromancer’s responsibilities went beyond ritual gatekeeping between the realms of the living and the dead. The necromancer, either male or female, was rather the direct intermediary, without any need for skulls or figurines as portals for communication. In this minimalistic ceremony, the spirit would blend with the necromancer, which could elicit the impression that the necromancer had become the spirit, temporarily at least. This kind of necromancy, to be distinguished from possession by malevolent spirits, might have been reminiscent of ancient and more recent forms of mediumship or, more broadly, mediumism (Paper 2004, pp. 41–45; Roxburgh and Roe 2013, pp. 60–64), in which the spirit communicator ‘overshadows’ the medium. The following quotation (Beard 1966, p. 104) describes the physiological aspect of mental mediumship as practiced within the Spiritualist Movement:

Communicators who resume their earth memories in order to give a message sometimes seem to be overtaken, as it were, by these memories so that in some compulsive way they find themselves temporarily reliving them. In particular they may resume the symptoms of their final illness and transfer these momentarily to the medium, who finds herself, perhaps, gasping for breath, or feels temporarily paralysed, or experiences an acute pain in the heart or chest. This is a very common phenomenon.

Back in Mesopotamia, this fleeting blending of the spirit and the divinatory specialist had its linguistic reflection in the term eṭemmu; depending on the context, it could mean ‘spirit of the dead’ or, secondarily, ‘necromancer’. Alternatively, rather than being a praxis of a bygone era, this more sober ceremony may have been a contemporary, affordable type of necromancy, which was practiced in the popular sphere.18 Hence, the term eṭemmu would have carried the secondary meaning ‘necromancer’ in everyday speech. Depending on the context in which the term was used, the people distinguished between its various meanings.

As speculative as the above suggestion seems at first glance, a similar semantic phenomenon is attested in another ancient Semitic language. In the Hebrew Bible, the frequently paired terms ’ôb (אוֺב) and yidd^e‘ōnî (יִדְּעֹנִי) refer not only to the spirits of the dead, or their images, but also to the practitioners who communicated with them.19 Although the practitioner interpretation for either term has at times been overlooked or dismissed in the scholarly debate (cf. Finkel 2021, pp. 252–54; Schmidt 1996, pp. 152–54; Tropper 1989, pp. 170–204), Hamori argues persuasively that it at least applies to 1 Sam. 28:3, 9 and 1 Chron. 10:13, and she renders these terms in this sense as ‘ghost-diviner’ and ‘spirit-diviner’, respectively (Hamori 2015, pp. 105–30; cf. Jeffers 1996, pp. 170–71). Moreover, various modern-day English Bible translations tend to apply the practitioner interpretation to other verses in which these terms are attested:20

	Koren Bible21	NKJV22	JPS Tanakh23	NRSV24
Lev. 19:31 ’ōbôt and yidde‘ōnîm	‘mediums’ and ‘wizards’	‘mediums’ and ‘familiar spirits’	‘ghosts’ and ‘familiar spirits’	‘mediums’ and ‘wizards’
Lev. 20:6 ’ōbôt and yidde‘ōnîm	‘mediums’ and ‘wizards’	‘mediums’ and ‘familiar spirits’	‘ghosts’ and ‘familiar ‘spirits’	‘mediums’ and ‘wizards’
Lev. 20:27 ’ôb and yidde‘ōnî	‘medium’ and ‘wizard’	‘medium’ and ‘familiar spirit’	‘ghost’ and ‘familiar spirit’	‘medium’ and ‘wizard’
Deut. 18:11 ̉’ôb and yidde̒ōnî	‘medium’ and ‘wizard’	‘medium’ and ‘spiritist’	‘ghosts’ and ‘familiar spirits’	‘ghosts’ and ‘spirits’
1 Sam. 28:3 ’ōbôt and yidde‘ōnîm	‘mediums’ and ‘wizards’	‘mediums’ and ‘spiritists’	‘ghosts’ and ‘familiar spirits’	‘mediums’ and ‘wizards’
1 Sam. 28:7 ba‘alat-’ôb	‘a woman who is a medium’	‘a woman who is a medium’	‘a woman who consults ghosts’	‘a woman who is a medium’
1 Sam. 28:8 ’ôb	‘familiar spirit’	‘séance’ (Hebr. ‘divine for me by the ôb’)	‘ghost’	‘spirit’
1 Sam. 28:9 ’ōbôt and yidde‘ōnîm	‘diviners’ and ‘wizards’	‘mediums’ and ‘spiritists’	‘ghosts’ and ‘familiar spirits’	‘mediums’ and ‘wizards’
2 Kgs. 21:6 ’ôb and yidde‘ōnîm	‘mediums’ and ‘wizards’	‘spiritists’ and ‘mediums’	‘ghosts’ and ‘familiar spirits’	‘mediums’ and ‘wizards’
2 Kgs. 23:24 ’ōbôt and yidde‘ōnîm	‘mediums’ and ‘wizards’	‘mediums’ and ‘spiritists’	‘necromancers’ and ‘mediums’25	‘mediums’ and ‘wizards’
Isa. 8:19 ’ōbôt and yidde‘ōnîm	‘mediums’ and ‘wizards’	‘mediums’ and ‘wizards’	‘ghosts’ and ‘familiar spirits’	‘ghosts’ and ‘familiar ‘spirits’
Isa. 19:3 ’ōbôt and yidde‘ōnîm	‘mediums’ and ‘wizards’	‘mediums’ and ‘sorcerers’	‘ghosts’ and ‘familiar spirits’	‘ghosts’ and ‘familiar ‘spirits’
Isa. 29:4 ’ôb	‘medium’	‘medium’	‘ghost’	‘ghost’
1 Chron. 10:13 ’ôb	‘medium’	‘medium’	‘ghost’	‘medium’
2 Chron. 33:6 ’ôb and yidde‘ōnî	‘mediums’ and ‘wizards’	‘mediums’ and ‘spiritists’	‘ghosts’ and ‘familiar spirits’	‘mediums’ and ‘wizards’

From the above overview, we gather that more formal equivalent translations, such as the NKJV and the Koren Jerusalem Bible, tend to apply the practitioner interpretation to the terms ’ôb and yidd^e‘ōnî. Whereas their specific understanding of yidd^e‘ōnî as a magico-divinatory practitioner is equivocal—the term is variously rendered as ‘wizard’, ‘sorcerer’, ‘spiritist’, and ‘medium’—these translations generally interpret the term ’ôb in a necromantic sense as ‘medium’ or ‘spiritist’. The comparatively dynamic equivalent JPS Tanakh translates the plural forms of these terms only once as ‘necromancers’ and ‘mediums’ (2 Kgs. 23:24), accompanied by a footnote, which provides the translation with ‘ghosts’ and ‘familiar spirits’. The NRSV occupies a middle position, though it tends to favor the practitioner interpretation throughout its translation of these terms, thereby unwaveringly rendering them as ‘medium’ and ‘wizard’, respectively.

The classical Hebrew lexica display a varied picture, with infrequent attestations of the practitioner interpretation.

	’ôb	yidde‘ōnî
BDB	1. skin-bottle; 2. necromancer; 3. ghost; 4. necromancy	familiar spirit
HALOT	spirit of the dead	1. spirit of divination; 2. soothsayer
DCH	1. ghost; 2. medium, necromancer	1. familiar spirit; 2. medium, necromancer

The story about the Woman of Endor in 1 Samuel 28 might hold the clue as to the use of these semantically ambiguous Hebrew terms in ancient Israelite society. Within the direct speech that is embedded in this narrative, we encounter various terms for necromantic specialists:26

^{Verse 7} Then Saul said to his servants, ‘Seek for me a woman who is a mistress of the spirits of the dead [ba‘^alat-’ôb], so that I may go to her and inquire of her’. His servants said to him, ‘There is a woman who is a mistress of the spirits of the dead [ba‘^alat-’ôb] in Endor’. ^{Verse 8} So Saul disguised himself and put on other clothes. Then he went, he and two men with him. They came to the woman by night. And he said, ‘Divine for me by the spirit [’ôb] and bring up for me the one whom I name to you’. ^{Verse 9} The woman said to him, ‘Look, you know what Saul has done, that he has cut off the mediums [’ōbôt] and the necromancers [yidd^e‘ōnîm] from the land. Why then are you entrapping my life to bring about my death?’

Interestingly, whereas King Saul uses the designation ba‘^alat-’ôb, ‘mistress of the spirits of the dead’, for a female necromancer, followed by his servants, the woman of Endor refers to her fellow necromantic specialists as ’ōbôt and yidd^e‘ōnîm, terms which, incidentally, can apply to both male and female practitioners (cf. Hamori 2015, p. 109). The king uses the term ’ôb solely in its primary sense as ‘spirit (of the dead)’, as is further evidenced in verse 8: ‘divine for me by the spirit [’ôb]’. The woman, on the other hand, employs the same term in its secondary meaning.

In my view, this difference in jargon might reflect a deliberate linguistic strategy: the composer, or editor, sought to mimic the sociolects of the characters involved in the story. King Saul and his servants converse at a higher register than the woman, who employs a rather more vernacular type of language.27 The colloquial terminology is adopted by the narrator of the story:

^{Verse 3} Now Samuel had died, and all Israel mourned for him and buried him in Ramah, in his city. As for Saul, he removed the mediums [’ōbôt] and the necromancers [yidd^e‘ōnîm] from the land.28

Hamori (2015, p. 106) regards ba‘^alat-’ôb as analogous to Akkadian ša eṭemmi, the professional necromantic designation which we encountered earlier in the present study. Notably, just as we search in vain for evidence of ša eṭemmi beyond the Mesopotamian lexical lists, there is no further attestation of ba‘^alat-’ôb, or its masculine counterpart ba‘al-’ôb, in the Hebrew Bible. If ’ōbôt and yidd^e‘ōnîm were colloquialisms for necromantic professionals, we understand their use throughout the Hebrew Bible, at least in the instances where the practitioner interpretation applies to them. Necromantic specialists were denounced in pro-Yahwistic prophetic, priestly, and scribal circles, which gained increasing influence in the final stages of the Judahite kingdom and the ensuing Babylonian exile. Given their fierce stance against necromancy, the composers and editors may have sought to address their audiences in the vernacular. These terms, which frequently feature within a legislative framework and prophetic discourse, would have resonated with common people. Depending on the context, they would have distinguished between the various meanings of ’ōbôt and yidd^e‘ōnîm.

Alas, the story about the Woman of Endor does not throw light on the actual necromantic practice of this ba‘^alat-’ôb. It is unknown whether she would have used incantations and ritual appliances, such as skulls and figurines, like the Mesopotamian necromancers from the mid-first millennium BCE, albeit in a less elitist setting. The story conveys the impression that Saul was able to converse directly with Samuel’s spirit (vv. 15–19), but it leaves us in the dark as to whether the woman functioned as a medium, in the modern-day sense of the word, or as a ritual gatekeeper whose main responsibility was the conjuration and controlling of the spirit, like the aforementioned ‘Totenbeschwörer’. Nevertheless, the semantic range of ’ōbôt and yidd^e‘ōnîm hints at a form of necromancy in ancient Israel in which the practitioner was the direct communicative channel between the living and the dead. The legislation in Lev. 20:27 might also refer to this mediumistic type of necromancy:

A man or a woman in whom there is a spirit (’ôb) or a ghost (yidd^e‘ōnî) shall surely be put to death; they shall be stoned to death; their blood is upon them.

The similarities between eṭemmu on the one hand and ’ōb and yidd^e‘ōnî on the other rest solely on the shared semantics between these three terms. They are not in the slightest etymologically related to one another. However, First Isaiah’s oracle against Egypt preserves a hapax legomenon that has been traced back to Akkadian eṭemmu. According to Isa. 19:3b, the Egyptians

will consult the idols [’^elîlîm] and the spirits of the dead [’iṭṭîm],

and the mediums (’ōbôt) and the necromancers (yidd^e‘ōnîm).

The practitioner interpretation seems befitting here for ’ōbôt and yidd^e‘ōnîm; in their despair and panic, the Egyptians will resort to necromantic diviners.29 Because of this lexical and semantic pairing in Isa. 19:3bβ, it appears that the terms ’^elîlîm and ’iṭṭîm in the preceding line have been used in a different sense. Scholars have understood the uniquely attested plural noun ’iṭṭîm as either derived from or cognate to Akkadian eṭemmū and translated it accordingly as ‘spirits of the dead’ (e.g., Schmidt 1996, pp. 154–58; Tropper 1989, pp. 278–84). Mankowski, however, cautions against a too hasty identification of ’iṭṭîm as a loanword derived from Akkadian eṭemmu (Mankowski 2000, p. 32):

‘The difficulty with the identification of אִטִּים and eṭimmu is twofold. On one hand, the spelling of the word and the parallelism of the verse suggest that אִטִּים was understood as a plural noun and its mem as the plural morpheme, whereas the -m(m) of eṭimmu is part of the root. On the other hand, the doubled second radical of a foreign borrowing usually points to transmission into Hebrew via Aramaic, in which the word is not attested. While the lexical appropriateness of eṭimmu for this verse seems too good to discount, we preserve the equation at the price of two ad hoc conjectures: the mistaken analysis (and re-patterning) of a singular noun as a plural, and the phonologically unmotivated gemination of its second root consonant’.

Bearing these caveats in mind, we continue our exploration of the link between ’iṭṭîm and eṭemmu. If there was indeed a close etymological relationship between these two terms, ’iṭṭîm might have also shared the wider semantic range that we established for its Akkadian counterpart. Interestingly, the practitioner interpretation has been applied to ’iṭṭîm in various Bible translations, lexica, and scholarly studies (e.g., Childs 2001, p. 139; Jeffers 1996, p. 167).

	Koren Bible	NKJV	JPS Tanakh	NRSV
Isa. 19:3 ’iṭṭîm	‘necromancers’	‘charmers’	‘shades’	‘spirits of the dead’

The Koren Bible and the NKJV interpret ’iṭṭîm as ‘necromancers’ and ‘charmers’, respectively. Henceforth, in both translations, the reference to the idols is followed by three categories of magico-divinatory specialists. The Brown–Driver–Briggs dictionary does not link ’iṭṭîm with Akkadian eṭemmu but rather traces the noun to the verbal root √’ṭṭ, of which the Arabic cognate has the meaning ‘to groan’. Consequently, this lexicon associates ’iṭṭîm with ‘mutterers’.

	’iṭṭîm
BDB	mutterers, i.e., either ventriloquists or whisperers of charms
HALOT	spirits of dead persons
DCH	ghosts

However, given the type of parallelism that characterizes Isa. 19:3b ’iṭṭîm has most likely been employed here in its primary meaning as ‘spirits of the dead’, thus forming a lexical pair with ’^elîlîm, ‘idols’.

5. Concluding Observations

Whether Imdī-ilum immediately gave heed to the alarming message sent by Tarām-Kūbi and Šīmat-Aššur remains unknown. It appears that the once successful merchant returned to Aššur later in life in an impoverished state after his business had been struck by misfortune (Hecker 2007, p. 93). Imdī-ilum had been forewarned: his love for money would lead to his demise. The sisters’ concern for their brother is remarkable; they resorted to three types of divination to learn the will of the god Aššur. Their reference to the consultation of the eṭemmū poses a fascinating necromantic riddle, the possible solution of which has led us on an extensive journey across space and time, from the homes in Aššur that simultaneously served as burial sites, via Ashurbanipal’s library in Nineveh, to the ancient Israelite town of Endor, which formed the backdrop of a secretive, nocturnal visit of King Saul to a female necromancer.30 Our exploration of the textual materials has yielded possible traces of a mediumistic kind of necromancy that left its imprint on the vernaculars of both Akkadian and Hebrew. Common terms for the spirits of the dead, such as eṭemmū, ’ōbôt, yidd^e‘ōnîm, and, to a lesser degree, ’iṭṭîm, may have served secondarily as colloquialisms for necromancers.

The final question to be discussed is whether this parallel semantic phenomenon in Akkadian and Hebrew was born out of language contact or a separate linguistic development. It is worth noting Schmidt’s thesis that necromantic practices only permeated Israelite religion under Neo-Assyrian influence during the reign of King Manasseh of Judah (Schmidt 1996, pp. 241–42). He finds insufficient proof for necromancy in Canaanite–Israelite religions before Manasseh’s reign.31 The following arguments have been brought forward by him in support of his theory:

‘(1) the late compositional histories of the relevant biblical texts as well as their traditions, (2) the rise in popularity of various forms of divination among late Assyrian kings, (3) the preponderance of references to Mesopotamian necromancy from the Neo-Assyrian period onwards, (4) the political domination of Judah by the Mesopotamian imperial states of Assyria and Babylonia in the mid first millennium, and (5) the evidence for Mesopotamian influence on the religious life of late pre-exilic Judah—whether by means of willful adoption or imperial imposition’ (Schmidt 1996, p. 241).

If we were to build on Schmidt’s thesis, we could argue that the colloquial use of the Akkadian term eṭemmu in the sense of necromancer may have influenced the semantic range of Hebrew necromantic terminology in late pre-exilic Judah. However, we should bear in mind that the necromantic manuals attested from the Neo-Assyrian period onwards are extensively ritualistic, learned texts, the consultation whereof was restricted to divinatory specialists who served the upper social classes. Thus, it seems doubtful whether such highly sophisticated esoteric knowledge would have circulated beyond the confines of the palace and temple and had a profound impact on ancient Israelite religion. Moreover, in these manuals, the term eṭemmu (GIDIM) carries its primary meaning as ‘spirit of the dead’, which further diminishes the possibility of Akkadian influence on the semantic range of Hebrew necromantic terminology.

Alternatively, this parallel semantic phenomenon may have developed independently in both languages under influence of a more common, simplified kind of necromancy, in which the spirit communicator briefly blended with the necromancer. This mediumistic type of necromancy, of which we find attestations through the ages and across various societies (Paper 2004, pp. 41–45), may have been indigenous to ancient Near Eastern folk religions. Modern-day mediums frequently recount how they were already attuned to the realm of spirit from a very tender age, passing on messages from the departed to the living, whether welcome or not (Roxburgh and Roe 2013, pp. 53–67). Similarly, back in Mesopotamia and ancient Israel, people with inherent mediumistic abilities may have wished to serve as intermediaries between the living and the dead, imparting knowledge that was otherwise hidden from mortals. This type of necromancy most likely required specific precautions as well, perhaps by means of incantations, prayers, and simplified rites, because of the capricious nature of the spirits and the possibility that eerie, malevolent forces awaited their chance to create havoc among the living.

In contrast to modern-day mediumship, as practiced for instance in the Spiritualist Movement, this necromantic praxis seems to have been predominantly focused on the hidden knowledge and intercession that the departed could offer. The evidential aspect occupied a less prominent role, most probably because the spirits of the dead were literally and figuratively ever present in people’s lives. This explains not only why Tarām-Kūbi and Šīmat-Aššur do not share with Imdī-ilum which deceased relatives had been consulted and any further information about them, but also the brief comment of the ba‘^alat-’ôb on Samuel’s appearance. When King Saul asks her what the spirit looks like, she simply states that she sees an old man coming up, wrapped in a robe (v. 14). However, this rather meagre clue is sufficient for Saul, who immediately pays obeisance to Samuel.32 Saul may have actually bowed down before the female necromancer, who not only imparted the prophet’s message but temporarily embodied him. As such, the ba‘^alat-’ôb had briefly become the ’ôb herself. Likewise, the ša eṭemmi could be overshadowed by the eṭemmu, a phenomenon that resulted in the wider semantic range of the latter term.