Some Remarks on Language Usage in Late Babylonian Letters

Abstract This paper deals with language usage in private and institutional letters from the Neo-Babylonian and Achaemenid periods. It analyses the usage of terms of address, greeting formulae and direct and indirect phrasing, drawing on notions of politeness developed by Brown and Levinson. Of particular interest are questions of usage within a temple hierarchy and its implications for professional relationships. For private letters, the case for the appellation of ‘lord’ for women as previously claimed by the author is further substantiated.

Babylonian period2 -have already emphasized that these letters are written texts and follow rules specific to the genre they belong to. For the corpus of Late Babylonian letters, we have almost 1.500 letters and letter orders3 in total (Hackl 2007:4-6).4 Of these, roughly 300 letters are of private origin (see most recently Hackl et al. 2014), from family archives, i.e. from archives of families which usually belonged to the upper and middle strata of society and which often had ties to larger institutions like the crown or temples.5 'Institutional' letters originate in an institution such as, in our case, temple archives. 6 For the purposes of our research, the main obstacles of working with these texts are the following: -the absence of contextualization: letters usually do not contain dates or affiliations.7 They mostly stem from illegal excavations and found their way to various museums around the globe via the trade in artefacts. -the low number of texts written by one person, especially in private contexts; -the uncertainties related to the process of letter writing, i.e. the problem of the possible involvement of a scribe and, in turn, possible consequences for language choice.8 In this paper, I am going to discuss three cases studies with a focus on different aspects of language use in different contexts: greeting formulae, directness and indirectness, and terms of address in private and institutional settings. I build on the notion of two poles of language usage, directness and indirectness, as heuristic markers for relationships.9 By applying it to administrative texts, I want to clarify language usage in more detail, especially whether personal bonds or professional rank were determining language choices on different hierarchical levels. Additionally, I will apply the framework of Brown and Levinson's (1987) politeness theory in detail to one passage of text to show that complex forms of politeness existed in Late Babylonian letters. In the case of private letters, I further substantiate the claim of Schmidl (2014:55-58) that women could be called 'lord'. With these goals in mind, I will refrain for the most part from treating the philological minutiae on the texts presented. Examples are given in translation and a pertinent transliteration can be found in the appendix.10

Methodology
Language usage in letters is reviewed here based on the politeness theory of Brown and Levinson (1987). They base their model of politeness on notions of "face". Every person has face, and they further distinguish between a positive and a negative face. Inherent in these are positive and negative face wants, with the positive ones depending on the need to be appreciated or acknowledged, and the negative ones being based on the need to be free and unimpeded in one's actions. It is assumed that based on rationality, the maintenance of face on all sides is in the interest of all parties. This is necessary because some speech acts inherently threaten face, so-called "face-threatening acts" (FTAs) (Brown and Levinson 1987:60). This means that face can be lost or saved. Based on this concept, they define hierarchically structured strategies for doing an FTA, with more indirectness used for higher-ranking strategies. If employed in the right context, indirectness bears less risk of an FTA failing. One can do an FTA "bald on record" (passim, esp. pp. 94-101), i.e., without redress, or employ redress in the form of positive or negative politeness (addressed to positive or negative face, respectively). Another possibility would be to go off record with an FTA ("more than one unambiguously attributable intention so that the actor cannot be held to have committed himself to one particular intent", p. 69) or, to refrain from doing an FTA altogether. The choice between these strategies is dependent on the parameters of "(i) the 'social distance' (D) of S and H (a symmetric relation) (ii) the relative 'power' (P) of S and H (an asymmetric relation) (iii) the absolute ranking (R) of impositions in the particular culture."11 (p. 74). This theory has been applied to different languages and has yielded positive results, but some aspects have been criticised. One problem treated e.g. by Ide (1989) is the supposed Eurocentrism of the theory, especially in the conception of face (see also Matsumoto 1988), which she sees as too individualistic for Japanese politeness. This notion has not remained uncontested (see e.g. Fukada andAsato 2004, Kiyama et al. 2012), but admittedly, there is the risk of inserting too much individuality retrospectively when dealing with a pre-modern society. It is impossible to know which precise conception of face is manifest in Late Babylonian letters, but the basic concept has successfully been applied to Old Babylonian letters by Sallaberger (1999) and to Late Babylonian private letters by  and will therefore be pursued further here.

Case studies
For the practical application of the model of Brown and Levinson, three different cases will be examined. I will consider different appellations, the question of directness and indirectness and the system of greeting formulae, drawing on private and institutional letters. Directness and indirectness in the letters is understood in a rather literal way and will be treated along the lines of grammatical items: verbal forms (i.e., imperative, prohibitive and precative12), suffixes (with the second person being more direct) and appellations (indirect; as opposed to usage of the second person only). The institutional case studies are taken from the letters of the Eanna temple in Uruk dating to the final decades of the sixth century, a corpus which comprises around 190 texts.13 Text numbers refer to published copies of the various letters in cuneiform.14 When dealing with private letters, the numbers of these letters in the new edition in Hackl et al. (2014) will be given in brackets in addition to the normal text reference, e.g. CT 22,6 (no. 22), to make them easily accessible in transliteration and translation.
A close analysis of some formal criteria of letters will elucidate language usage in an administrative setting to see if personal or hierarchical elements can be made out to be determining factors. While this is an important goal in and of itself, it is also indicative for the bureaucratic mentalities of these officials, i.e. whether the administration was more patrimonial or rational in a Weberian sense.15 As a hypothesis, I assume that consistent usage of forms adhering more to the level of power instead of social closeness (i.e., based on offices, not persons) would mark "formality" in an administrative setting.16

The case of Nabû-aḫu-iddin, temple supervisor
Nabû-aḫu-iddin is one of the highest officials in the temple of Eanna. His official title is 'supervisor of Eanna' (ša rēš šarri bēl piqitti ajakki) (see Kleber 2008:12-27 for an extensive treatment). He was in office from 17 Nbn until 4 Camb (Kleber 2008:36). In this position, he shared his responsibilities with another official, the 'temple administrator' (šatammu). The supervisor of Eanna had a background in the royal administration, while the temple administrator stemmed from local families. The interests of crown and temple -and the powerful families connected to them -were not always in agreement and could lead to conflict. This means that although the two officials shared the same responsibilities, they often had different priorities. Nabûaḫu-iddin's letters provide an opportunity to analyse this tension in more detail.
In total, sixteen letters are preserved which were sent by Nabû-aḫu-iddin.17 Eleven of these are addressed to a temple administrator18 and four to a temple scribe19 (in YOS 21, 72, the addressee is broken20). This dossier is useful because it consists of letters addressed to offices with clearly different hierarchical status, with the temple scribe being of lower rank. One of these scribes, Nādin, is addressed in various letters (see also the next case study below). He seems to have been in charge when neither the temple administrator nor the supervisor of Eanna were present at the temple (Kleber 2008:28).
At the outset, we cannot discern these two groups of letters by appellations in the heading. All of his letters are addressed to a 'brother' (aḫu), i.e. to people of equal or lower standing, as expected. Nonetheless, there are distinctions between letters addressed to these two offices.21 One element is the greeting formula. All his letters addressed to the temple administrator employ a greeting wishing well-being (šulmu [u balāṭu] … qabû). In letters addressed to Nādin, the temple scribe, he can do this as well (YOS 3, 33 and BIN 1, 7), but he can also use the blessing formula (containing a form of karābu) (YOS 3, 85; TCL 9, 129).22 Frahm and Jursa (2011:8) deem the blessing formula to be "particularly appropriate for addressing a person of equal or inferior standing" based on their survey of its combination with appellations in the sample of Eanna letters published in YOS 21, where it refers to a 'lord' (bēlu) only once. It can be used like this in the early institutional letter corpus, but also in e.g. letters to the king, which are unambiguously 15 Weber (198015 Weber ( [1922) distinguishes between these two terms as ideal, opposed types of administration. They are never found in reality, but are useful as heuristic tools. 16 For Late Babylonian letters, Jursa (2014:78) showed that the nature of the exchange of messages was more formalised in the context of temples than in private correspondence based on the relatively higher weight placed on letters -as opposed to messengers alone -in the temples. For a general treatment of the various notions of formality in ethnographic research, with a special focus on communication, see Irvine (1979). 17 At least some of these are likely to have been written by himself, see Jursa 2012:380 andLevavi andSchmidl 2016. 18 YOS 21, 69 (using a personal name, not a title), 70 and 71;YOS 3,17,19,42,66,69,79 and 107;TCL 9,98. 19 YOS 3,33 and 85;TCL 9,129;BIN 1,7. 20 Judging from the content, it could be addressed to either Nādin, a temple scribe (see immediately below), or the temple administrator, but since we have more letters addressed to the latter, this seems more probable. It consistently employs direct address, unfortunately not a distinct characteristic of the letters addressed to one of these officials (see below). 21 For a few examples of rhetorical strategies Nabû-aḫu-iddin employs when writing to the temple administrator, see Kleber 2012b:227-228. 22 The different types of possible greeting formulae have been treated by Schmidl (2014:10-16) for private letters and by Levavi (2016:46-52) for early institutional letters (i.e., from the beginning of the Neo-Babylonian period until ca. the middle of the reign of Nebuchadnezzar II, 626 until ca. 585 BCE). A collection of all attested variants of various periods can be found in Salonen 1967. bottom-up.23 In private letters, the blessing formula is mainly employed in letters between equals or bottomto-top, but it can occur in top-to-bottom letters (Schmidl 2014:13). It is of course not possible to deduce a general rule from Nabû-aḫu-iddin's usage, but it is noteworthy that despite the much higher numbers of letters to the temple administrator, he does not employ the blessing as a greeting there. Whether this is due to personal choice only or part of a more general tendency will require further research on the usage of greeting formulae in the late corpus.24 Another distinction is the usage of direct and indirect address. Direct phrasing (i.e., second person and imperative) is employed in letters addressed to the lower-ranking temple scribe, while in letters sent to the temple administrator, both can occur. The most indirect phrasing, the appellation 'lord' combined with a third person precative, only appears in letters to the temple administrator. Still, it is not obligatory: direct address can be employed in other letters to the temple administrator and both forms of address can be mixed even within a single letter. The first, direct, usage is congruent with the expected difference in hierarchy, while the second, more varied usage can be seen as an expression of the shared responsibilities and equal power of the two offices.
The clearest example of this distinction can be shown by a comparison of YOS 3, 17 and its near duplicate, TCL 9, 129, written on the same occasion. The first one is addressed to the temple administrator, the second one to the temple scribe, employing different greeting formulae. Moreover, in l. 5 (YOS 3, 17) or 6 (TCL 9, 129), we observe a difference in directness and indirectness although the rest of the sentence employs imperatives in both letters. In the letter addressed to the temple administrator, Nabû-aḫu-iddin uses an appellation and a precative in the third person ('the lord imposes'), while the same issue is expressed directly in the letter to the temple scribe ('you impose'). Interestingly, other verbal forms are phrased directly.
(2) May Nabû and Marduk bless my brother. (3-5) As one can see, I have sent Šulāya, Nādin, Naˀid-Ištar, Mukkēa and Šadūnu to you. (5) Balance the account with them! (6-7) Receive from them whatever you impose upon them (as payment) and send them (back) to me quickly! (TCL 9, 129: 1-7) (Appendix no. 2). This distinction can be observed in more detail in Nabû-aḫu-iddin's usage of a stock phrase on negligence ('do not be negligent concerning x,' ana/ina muḫḫi x lā + form of šelû). It can be understood as an FTA imposed on the other person, i.e. threatening negative face. In letters addressed to the temple scribe, he uses the idiom consistently in the second person ('do not be negligent', la tašelli) (YOS 3, 85:11; TCL 9, 129:41), while he always phrases it indirectly ('the lord should not be negligent', bēlu la išelli) in letters to his equal. This is not dependent on the rest (or parts) of his letter being written in the second or third person (YOS 3, 79:20, see below; YOS 3, 66:27; YOS 3, 19:31(?) [see note 58]). Taken at face value as a heuristic marker, indirectness in this stock phrase would be employed for people of higher standing -as is done in private letters (Schmidl 2014:39). This could then be read as a sign for the superiority of the temple administrator in their supposedly equal relationship. But if we look at it from the point of Brown and Levinson's politeness theory, one can also suggest that Nabû-aḫu-iddin felt that this stock phrase was more of a threat to the 'face' of the temple administrator compared to other requests. This would make it necessary to use it indirectly even in situations where other verbal forms were used directly. In line with this observation, this stock 23 Levavi (2016:50) does not state this explicitly, but it is implicit in his mentioning of the two appellations used with this formula, 'brother' and 'lord', and can easily be extracted from his text editions, see e.g. YOS 3, 7 (Levavi 2016, no. 57) and GC 2, 395 (Levavi 2016, no. 67) for letters addressed to the king vs. examples of letters sent by the highest officials to people of lower standing, e.g. YOS 21, 127 (Levavi 2016, no. 131) or YOS 21, 50 (Levavi 2016, no. 134). 24 Already Levavi (2016:50) suggests personal preference as a possibility when it comes to the choice of greeting formulae. 25 Different gods and goddesses can be invoked in the various greeting formulae. They are not dependent on the formula, but often reflect a personal choice. This can be of use for the attribution of letters to various archives if this choice contains a local component. For an overview of the entities invoked in greeting formulae of the early Late Babylonian letter corpus, see Levavi 2016:46-48. See also Frahm and Jursa 2011:8-9.
This latter interpretation fits usage in early institutional letters, but not in private ones. For private letters, I suggested that this stock phrase was more polite than another one emphasising urgency (lit. 'not to stay the night', i.e., 'do not waste time', nubattu lā + form of bâtu), since the latter is only rarely aimed at higher-ups (Schmidl 2014:38-39). In early institutional letters, the phrase of urgency is used in letters from higher-ups and between equals (Levavi 2016:61). In line with this, Nabû-aḫu-iddin uses it in the second person in letters to the temple scribe (YOS 3,33) and to the temple administrator (YOS 3, 19,26 see below).27 Observations from private letters can therefore not simply be extended to the institutional corpus. But I will return to this stock phrase in section 3.2.
(1-2) Letter of Nabû-aḫu-iddin to Nādin (temple scribe), my brother.  If we want to take this as an expression of the power difference between these two supposed equals, we need to consider other verbal forms as well. As already shown by the previous examples, usage is by no means unambiguous and can change even within a single letter. For a better understanding of this phenomenon, one can again draw on the parameters of Brown and Levinson: in this case, the quality of the personal relationship. We know that there were three people in office as temple administrators during Nabû-aḫu-iddin's term of office. Their different relationships to him are likely to have influenced his choice of expression, but a comparison is difficult. One problem is that his letters usually address the temple administrator by title only. Kleber (2012a) overcame this obstacle and dated several of his letters to 2-3 Camb. Also Tolini (Tolini 2011a:153)28 identified three letters which were sent within a period of three months in 2 Camb, YOS 3, 79 and YOS 21, 70 and YOS 3, 19. Through prosopography, we can identify the person in office back then: Nabû-mukīn-apli (Kleber 2008:34). If we look at these three letters as an example, we see varied usage of directness and indirectness. Among these, YOS 21, 70 is the most indirectly phrased letter, inter alia with the appellation 'lord' used without a possessive suffix, i.e. most distant. YOS 3, 19 employs a mixture of both and YOS 3, 79 is the most direct example (except for the aforementioned stock phrase of negligence). Usage was therefore not dependent on the office of temple administrator alone. It is more likely that direct and indirect address depended on the context of a request. With Brown and Levinson, we can discern different politeness strategies, with direct address appealing to the recipient's positive face in some instances, and indirectness addressed to the recipient's negative face in others.29 26 In this letter, he also uses la išelli, but it is not sure whether he addresses this to a 'lord' or to 'someone', see note 58. 27 YOS 3, 39, uses it in the second person as well, the letter is addressed to a 'brother' and is probably exchanged between lower temple officials. Other instances in the late Eanna corpus are referring to a third person altogether, e.g. YOS 3, 26, 40 and 111. 28 See also Kleber 2008:188-191 andKleber 2012a for connections between his letters and their dating. 29 In line with Brown and Levinson (1987), one could suggest that direct usage could be a case of the mode "bald on record", which can be a sign of urgency or the need for efficient communication. Brown and Levinson (1987: 69) give three examples of possible circumstances for leaving out any redress of the fact-threatening act: "(a) S and H both tacitly agree that the relevance of face demands may be suspended in the interests of urgency or efficiency; (b) where the danger to H's face is very small, as in offers, requests, suggestions that are clearly in H's interest and do not require great sacrifices of S (e.g. 'Come in' or 'Do sit down'); and (c) where S is vastly superior in power to H, or can enlist audience support to destroy H's face without losing his own." (emphasis in the original). That the progress of the building projects was in the interest of the temple is clear, but it might still have required great effort to supply the goods and workers needed to achieve this, which makes this possibility unlikely.
(6-7) Concerning the dates which I wrote to you about (…) (12) I want to hear your opinion on this. (13-15) The message (…) which he wrote to you -it's indeed very good (news). (15-16) Have everything he wrote to you for (=he needs from you) delivered to him. (17-20) Concerning (…) and the needed supplies -the lord should not be negligent (about them). (20-21) Have beer delivered quickly and your messenger (…) (29) I did not see the silver which you had delivered (to me). (30) How much is it? (30-32) If you (already) had ten minas of silver delivered, add (more to it) and have (another) ten minas of silver delivered! (33) My messenger should not stay with you. (YOS 3,(12)(13)(14)(15)(16)(17)(19)(20)(21)(29)(30)(31)(32)(33)  For a comparison of the role of personal relationships, we unfortunately have only one other letter definitely addressed to a different temple administrator, Nidinti-Bēl (YOS 21, 69). We know this because he addresses him by first name, which is unique in his letters to the temple administrator. The letter is rather short and phrased directly. In this case, usage of the first name instead of an official title fits the directness of the rest of the letter. Levavi (2016:44) noted that usage of the first name instead of a title "reflects a certain degree of familiarity, though not necessarily a close personal relationship". Whether this was true for Nabû-aḫuiddin's general relationship with Nidinti-Bēl or only invoked in this letter to address Nidinti-Bēl's positive face remains open.32 In any case, this sample is too small to draw conclusions on Nabû-aḫu-iddin's language usage when writing to different office holders.
Comparing usage in letters to the two different offices, it is apparent that Nabû-aḫu-iddin's clear, hierarchical relationship with the temple scribe results in consistent, direct language usage. On the other hand, his ambiguous relationship with the office of the temple administrator manifests itself in a more complicated usage of directness and indirectness which should be located in the realm of politeness. We can also observe a possible distinction between appropriate greeting formulae to the various addressees, though, at this point, no definite rules can be deduced. The analytical framework presented here will be further expanded on in the next case study of a middle-ranking official.

Formal aspects of institutional letters: the case of Innin-aḫḫē-iddin
Innin-aḫḫē-iddin was a 'chief of serfs' (rab širkī) at the Eanna temple from 17 Nbn to 7 Camb (Kleber 2008:112). This official was chosen when necessary to oversee the Eanna's involvement in royal building projects (Kleber 2008:111). So far, eleven of his letters have been identified.33 In ten of them, he writes to the highest officials of the temple, the trustee (of Eanna) (qīpu), the temple administrator and the supervisor of Eanna, as well as the temple scribes. As the chief of serfs is positioned further down in the administrative hierarchy of the Eanna than these two groups, his choices in addressing them are limited to 'lord' or 'father' (abu).34 His dossier can enrich our picture of language usage from the perspective of a middle-ranking person.
Appellations and greeting formulae are analysed first. The heads of his letters present a clear picture: whenever he writes to the highest temple officials -from one only to all three of them -he consistently uses the appellation 'lord(s)'.35 Letters addressed to the temple scribes (one36 or several37) call them 'father(s)'.38 Moreover, he distinguishes between these two groups by greeting formulae as well. In letters addressed to the highest temple officials, the praying formula (containing a form of ṣullû) is used, while the wish for well-being is employed for temple scribes. The formulaic part of the letter therefore shows a clear picture of hierarchies and appropriate appellations congruent with different offices.39 In private letters, the address 'father' indicates a closer social relationship than 'lord' (Schmidl 2014:25). But Hackl and Jursa (2015:105-106) noted that 'father' is employed differently in institutional letters, where it expresses a stronger component of subordination,40 but it is used only for members of the sender's section of administration. Moreover, they suggest that its usage shows "strategic or persuasive intent" (p. 106), and in many cases implies a closer social relationship than 'lord', while 'lord' was appropriate for higher-ups of different administrative units. They do not explicitly distinguish between usage in the head and the body of letters. In our case, a more personal component of the appellation 'father' is unlikely with this clear distinction between hierarchical levels. Moreover, the high-ranking trustee (of Eanna) and the chief of temple serfs worked together on building projects (see Kleber 2008, 108-112, esp. 111). It remains to be seen whether this clear hierarchical distinction can be maintained for the later corpus in general. It is admittedly not observable in the early letter corpus. In fact, the appellations 'father' and 'lord' are once even used interchangeably between the same two parties (Levavi 2016:44). Usage in Innin-aḫḫē-iddin's case is more systematic. This could possibly reflect a developing administrative system with more rigid hierarchies.  While the opening formulae of the letters are schematic, usage in the body varies. Letters addressed to lords do not use the third person exclusively: this can be the case as in YOS 21, 35, to the supervisor of Eanna, but 34 These choices can be observed in early administrative letters (Levavi 2016:44). In private letters, we can see exceptions to this rule, where the possibility of using 'brother' as an appellation for higher-ups can be granted to people of lower standing (Schmidl 2014:22). 35 YOS 3, 10, 21, 45; YOS 21, 35. 36 YOS 3, 116, BIN 1, 29 and Stolper 2003, 272-274, all addressed to Nādin (see above, chapter 3.1). 37 YOS 3, 81 and BIN 1, 16. 38 There is one more letter addressed to a father, YOS 3, 106, but the recipient's name is broken. 39 Kleber (2012b:226-227), in her treatment of rhetorical strategies of temple officials, has shown that Innin-aḫḫē-iddin also uses different approaches in letters to the highest officials than in those to the temple scribes. 40 This subordination was not a one-way street, but was based on the notion of patronage, which was a reciprocal relationship, granting protection and support.

Unangemeldet
Heruntergeladen am | 09.11.18 10:00 usage can also be variable, see e.g. YOS 3, 21, to the temple administrator and the supervisor of Eanna.42 It can even employ direct forms only, as in YOS 3, 10, addressed to all three high officials.
(8-11) As one can see, I am sending PN 1 , son of PN 2 , to my lord. (12)(13)15) Quickly, the lord should give him 500 (kurruunits) of barley (…). (YOS 21,15) (Appendix no. 12). The same is true for letters addressed to 'fathers'. BIN 1,29,addressed to Nādin, the temple scribe (see above, chapter 3.1), uses the appellation 'father' in combination with a precative in the third person, while BIN 1, 16 combines it with imperatives only. Usage in other letters varies between the second and third person: YOS 3, 81, addressed to four scribes (including Nādin), and YOS 3, 116, addressed to Nādin as well.43  Judging from the openings of the letters, the relationship between the overseer of serfs and the various officials was clear-cut on the surface of matters. But when we look at the body of his letters, we see that this was not the case in everyday communication and in every situation. Assuming that all direct usage by subordinates was per se impolite would go too far. A closer look at the dispersion of direct and indirect forms along the lines of politeness strategies, FTAs and contexts will be illuminating in this respect. I will investigate here one case only, a passage in YOS 3, 45.
In YOS 3, 45, Innin-aḫḫē-iddin addresses the temple administrator on the one hand indirectly as 'lord', combined with forms of the third person, and on the other hand directly, using forms of the second person and imperatives. Indirect usage is slightly more predominant. He employs four precatives in the third person combined with 'lord' and three additional instances of this appellation alone. Direct means of address are only slightly less common in this letter, with five verbal forms in the second person (two in the imperative).44 42 See also YOS 3, 45 to the temple administrator further below. 43 One letter addressed to a father uses direct forms only and no appellations (YOS 3, 106), but the recipient's name is broken. 44 The usage of the personal pronoun of the second person (attā) is difficult to fit into this picture. It has been taken to be of neutral meaning in private letters because the third person pronoun is unattested, see Schmidl (2014:19). In institutional letters, attā features even in direct combination with the appellation 'lord', see, e.g., BIN 1, 92 (no. 89 in Levavi 2016) or YOS 21, 172 (no. 29 in Levavi 2016 for early examples, but also in e.g. later letters from the times of Nbn, BIN 1, 55 (edited in Hackl et al. 2011:196-197, see p. 183 for its dating) and YOS 3, 190 (for more references see Waerzeggers 1997:31-32). For institutional letters of the early period, the third person pronoun šū is used only twice in this manner, in BIN 1, 43:6-7 (no. 143 in Levavi 2016) and BIN 1, 42:12 (no. 2 in Levavi 2016) though he translates this example differently due to context). So far, it is not attested in the late letter corpus. With this considerable bias towards the usage of the second person pronoun, one might suggest that its usage is to be interpreted as neutral, as was done for private letters.

Unangemeldet
Heruntergeladen am | 09.11.18 10:00 I will focus on the passage about rations, ll. 8-25. Within this passage, the writer employs a prohibitive in the second person in l. 21 ('do not be negligent'). Interestingly, this is one of the rare instances of the aforementioned stock phrase for negligence (chapter 3.1). Its direct usage stands diametrically opposed to usage by Nabû-aḫuiddin, who employed this phrase in the third person even when dealing with a person of equal standing. Comparison to more administrative instances of this stock phrase does not elucidate this matter further. There are only two more attestations of this phrase in the late letter corpus from Uruk, and they derive from letters which were sent by the temple administrator (YOS 3, 156) and the supervisor of Eanna (YOS 3, 85), respectively, and addressed to people of lower standing. This means that their usage in a direct form does not tell us much about the possible connotations of this phrase.45 The attestation in YOS 3, 45, on the other hand, fits usage in private letters, in which this stock phrase can -though rarely -be addressed to people of higher standing, and even in direct phrasing if the personal relationship between the parties was close (Schmidl 2014:38-39). Since we cannot see evidence of a close personal relationship in our case here, one solution would be to attribute direct or indirect usage to personal perceptions of the implications of this phrase -in this hypothesis, severity of the FTA -by the individual writers in the relevant context.
If we take this phrase not on its own, but consider it in the context of this letter, a further explanation for its direct usage presents itself. Already two sentences before this direct request, the second person is employed in l. 18 ('you gave me'), but following indirect usage in l. 12 ('My lord should consult ledgers'). It is thereby contrasted with the request that older records should be taken into account. The amount delivered in the old documents is linked to the amount Innin-aḫḫē-iddin has already received from the temple administrator. Directness here involves the temple administrator more closely and draws attention to the fact that he had already delivered goods according to the sender's expectations and was therefore personally involved. But this directness is not pursued immediately. Instead, the sender changes back to the third person directly afterwards in l. 19, 'the lord should make the expenses', a case of negative politeness. Then, in l. 21, the second person is employed via a prohibitive, 'do not be negligent', the stock phrase in question. I would suggest to see this as a change in politeness strategy, to positive politeness, which can be connected to notions of solidarity and appreciating the other's positive face. If this is indeed the case, the more direct form would express solidarity by signalling more closeness between the two parties involved, probably inferring that it was in the best interest of both letter parties that these rations were delivered and that they were both on the same team to ensure the success of this endeavour. Moreover, by not explicitly spelling out what needs to be done to comply to the request of non-negligence, the speaker involves the hearer by making him assume what the necessary steps are which need to be taken. This hints at commonality of the parties again, this time in knowledge.46 45 The few attestations of this phrase could hint at restricted usage -it might not have been possible for every sender to use this admonition -or any, for that matter -when addressing a recipient. 46 Staying with the model of Brown and Levinson, one could alternatively suggest that this prohibitive was not a strategy of positive politeness, but a request stated "bald on record", underlining the urgency and importance of this request and hinting at a common goal. This notion of urgency would then also be stressed by the impersonal 'the rations must not fall short' at the end of this passage. This is unlikely, since "bald on record" can be applied only in ratherlimited situations (see above, note 29), and would probably be imperatives exactly delineating what the recipient should do, e.g., get the rations, send them etc. Considering the overall mixed usage of direct and indirect forms in his letters, an explanation as a strategy of positive politeness is much more plausible.

Unangemeldet
Heruntergeladen am | 09.11.18 10:00 In the next passage, ll. 21-25, Innin-aḫḫē-iddin shows his willingness to do his part to keep the rations from falling short, though this is dependent on the addressee's actions, a delivery of goods. To request these goods, he employs a third person precative with an appellation in l. 23 ('my lord should have delivered'), returning to a form that shows distance again. I take these indirect passages to be a sort of frame for the more direct approach. I consider the ranking of impositions to be roughly the same in the context of the larger goal of delivering rations.47 The section closes in l. 25 with a general, impersonal statement that the rations are in danger of running out ('the rations must not fall short'), implying that this needs to be avoided at all costs,48 but without referring to either party of the letter.
Taken together, several possibilities to phrase requests are open to Innin-aḫḫē-iddin in the dynamic field of power and personal relationship between him and the recipient. He employs a dual strategy of directness and indirectness which goes beyond a simple and strict necessity to adhere to negative politeness as in the heads of his letters. I interpret the change in politeness strategies as a way of negotiating their relationship: the opening with indirect usage sets the stage and enables him to use positive politeness. The following recourse to a more distant form can be explained in two ways -as further emphasis on the usage of directness, or that it was necessary to embed the direct form in this indirect frame for it to be applicable in the first place. A direct approach to the recipient can stress common interests and sometimes in-group mentality or common ground, which fits the importance of this request for the temple as a whole. In the framework of Brown and Levinson (1987:103), directness can also function as a "social accelerator" signifying that the person using it wants to "'come closer'" (p. 103) to the recipient, which could also have been an issue here if Innin-aḫḫē-iddin desired a certain status or relationship with the addressee.49 Nonetheless, positive politeness is, on the scale of strategies presented by Brown and Levinson, in theory a more risky strategy -not employing the right level of politeness can cause a request to fail. Whether this change here was a conscious choice by the sender or not cannot be determined. Based on the assessment above, I suggest that Innin-aḫḫē-iddin's choices of directness and indirectness in this case are based on politeness, applying different forms of redress. This implies that -to a certain extent -free usage of direct and indirect forms was possible for people of lower rank as well and was not only applicable to equals as we saw in the first case (chapter 3.1).
To sum up, language usage in the heads of Innin-aḫḫē-iddin's letters consistently reflects the differences in the hierarchical positions of sender and addressee. This can be interpreted as a more systematic, probably more formal, usage in institutional letters addressed to higher-ups and could even indicate a form of "standardized" language usage. The appellation 'lord' in letters addressed to scribes would then be an optional polite form, but only in the body of the letters. In these less formalized parts of Innin-aḫḫē-iddin's letters, language usage is not as unequivocal as in the headings. It is possible for the chief of serfs to employ a wider range of politeness strategies than an assessment purely based on rank would have provided. While his uniform letter headings might indicate a higher degree of standardisation, the body of the letter shows that even in institutional settings and in letters to higher-ups, the linguistic choices could be broad and varied, and a personal component cannot be ruled out as an influencing factor.

Appellations in letters addressed to women
For the last case study, I will draw on the corpus of private letters. Institutional letters hardly ever address women since women did not hold office in the temple.50 Private letters addressed to women were treated by Schmidl (2014: 55-58). Letters addressed to women are rare and make up only around eleven percent of the letters published in Hackl et al. (2014) (Schmidl 2014:53). Next to the female appellations known from other letters ('lady', 'mother', 'sister'), which are paralleled by those used to address men, the address to women could also employ the appellation 'lord', though only in the body of the letter (for a detailed treatment, see Schmidl 2014:56-57). The best and unambiguous example is no. 92 in Hackl et al. (2014), where the appellation of 'sister' in the head of the letter changes to 'lord' in the body of the letter. The distinction between usage in the head from that in the body is reminiscent of the systematic character of the formal parts of the letters of Innin-aḫḫē-iddin and its contrast with the rest of the letter, where 'lord' is a possible polite variant for the temple scribes.
I would like to use this opportunity to corroborate further my previous statement that 'lord' is a possible appellation for women. The appellation 'lady' (bēltu) is attested in letters, which means that employing an appellation 'lord' for women is not a necessary or obvious choice. But if we look at the letters which employ the appellation 'lady' more closely, we see that none of these letters repeats the appellation 'lady' in the body of the letter. Instead, the second person is used in the form of a suffix51 or a prohibitive (CT 22,6 [no. 22]:21). A complete switch to the appellation 'lord' is also attested (CT 22,200 [no. 47]). This system is more reminiscent of the usage of 'father' than 'lord' in letters addressed to men.52 The female pendant to 'father', 'mother' (ummu), is only employed once in the whole corpus of private letters (YOS 3,22 [no. 232]). In this case, the appellation is employed consistently throughout the letter, using it to distinguish parts of the text addressed to a man. The letter was officially exchanged between a nephew and his aunt, but in some passages, he addresses his uncle as well. This necessitated a way to keep the two addressees apart. Since 'mother' seems to be an unusual choice and 'lady' is employed in a manner reminiscent of the usage of 'father', there is no option mirroring the appellation 'lord' in this framework, i.e., a distant appellation which is combined with the usage of third person forms in a majority of instances in private letters (Schmidl 2014:25). I therefore suggest that the appellation of women as 'lord' was introduced or at least used as a distinctly polite form for addressing female correspondents. This is congruent with the previous suggestion by Schmidl (2014:56-57) that 'lord' had to be a polite form of address for women, conveying great respect, likely because of a discrepancy in gender.
We can further substantiate the argument for the appellation of 'lord' for women by drawing on material from the Neo-Assyrian period, i.e. Neo-Assyrian and Neo-Babylonian letters. Svärd (2015:55-58, 82-83) has collected evidence on the use of titles and appellations for women.53 The appellations are taken from letters from royal archives, which address women of very high standing. Here, we can observe the appellation 'lady' in the head and body of letters, e.g. for Naqiˀa, queen and mother of King Esarhaddon (SAA 10, 154, Neo-Babylonian).54 On the other hand, it is also possible to address important women as 'lords' in the head and in the body of letters, also to the aforementioned Naqiˀa (e.g. SAA 18, 85, Neo-Babylonian).55 Interestingly, the only two letters which depict Naqiˀa in a governing position in Babylonia both use the appellation 'lord' (SAA 18, 10 and 85, Svärd 2015:57). Some instances in the heads are not fully conclusive: letters written to queen mothers open with the phrase "to the mother of the king, my lord" ('my lord' realized as bēliya, oblique case), where the appellation can apply to the king or his mother, i.e. "to the mother of the king, my lord," or "to the king's mother, my lord". Usually, the first interpretation is preferred, but Svärd (2015:83) shows that it can unambiguously stand for the king's mother as well, in cases where 'my lord' is realized as bēlī (nominative), and should therefore also be taken into account as a possibility for oblique cases. Moreover, this can be combined with masculine verbal forms.56 Also, oblique cases in the appellation 51 CT 22, 6 (no. 22):19, letter from son to mother, and YOS 21, 36 (no. 234):14, sender not identifiable with certainty. 52 For a detailed treatment of the usage of 'father' and 'lord' in private letters, see Schmidl 2014:18-20 and 25-26. 53 References given here are taken from Svärd (2015: 55-58, 82-83). See there for more references, also from a non-epistolographic source. 54 For more information on her, see e.g. Svärd 2015, 52-59, Macgregor 2012, 95-122 and Melville 1999 According to Svärd (2015:55-56), several letters from high-ranking people to Naqiˀa are phrased highly politely. She even received compliments usually reserved for the king (Parpola 1983:176, 220). 56 For references, see Svärd 2015:83 410 .
Royal correspondence is of course not directly comparable to private letters, especially from another period, but it does provide additional evidence on this subject. It illustrates that the appellation 'lord' was a valid possibility in Akkadian to address women, and that it was in use for high-standing women in the Neo-Assyrian period. Our case is an expression of a similar socio-linguistic phenomenon, but a direct connection can of course not be found -its origins can just as well be fully autonomous. In detail, usage is slightly different in Late Babylonian private letters, i.e., only in the body of a letter, though this could be due to the small sample.

Conclusions
This paper expands on previous applications of the framework of Brown and Levinson to Babylonian letters. The first two case studies show the possible applications and limitations of this approach for institutional letters. The underlying historical issue here is the nature of hierarchical structures within a temple, i.e., the importance of personal relationships vs. impersonal administrative structures. The investigation shows that only letters to people of lower rank are consistent in their reference to the objective hierarchical situation. In cases of letters to equals in rank, language usage is much more varied. The relationship between the two principal offices in question, those of the temple administrator and the supervisor of Eanna, cannot be clarified further on the formal grounds examined.
In letters from people of lower standing, appellations in the opening as well as a set number of greeting formulae are used systematically when addressing holders of offices of different (though always higher) status. These regularities are interpreted as more formal usage compared to the one in private letters. Nonetheless, this cannot be taken as a more general tendency toward impersonal administrative structures, as it is countered by language usage in the body of these letters. There, usage is not consistently indirect to express deference, but it is varied and allows for different politeness strategies.
The last case study provides further evidence for a previous finding in Schmidl (2014:55-58), i.e. that women could be addressed as 'lord'. Moreover, a certain degree of formalisation can be discerned in private letters as well. The more formal heading cannot include the most revering address of 'lord' for women, but it is possible in the body of the letter. Additionally, evidence from the Neo-Assyrian period shows a similar socio-linguistic phenomenon, though no direct link is postulated.

Abbreviations
Abbreviations of text numbers follow the conventions of Archiv für Orientforschung. Nbn Nabonidus (555-539) Cyr Cyrus (538-530) Camb Cambyses (529-522); e.g. 2 Camb equals parts of 528 and of 527 BCE. PN personal name pl. plural The texts in the appendix are arranged according to their position in the article.