With apologies to Lucy Van Pelt, I thought I made a mistake but I was wrong.
The borrowed writing system that scribes used for Akkadian can be very vexing. It alternatively provides less information than one would like and more information than one wants. The last word in one line of an Akkadian prayer I’m working on reads in my draft i-ši-ṭa-a[n-ni]. The last two signs are a partially reconstructed sign followed by a completely reconstructed sign. That’s what [] means. But the reconstruction is not the problem. It’s the clearly readable part of the text that is vexing. Note the dot under the t. If this reading is correct the infinitive of the verb is šêṭu. But when Alan Lenzi saw my draft he reasonably commented, “Check your text. I think the verb is šêtu.” And as soon as I saw his comment, I too thought it likely that the verb was šêtu and that I had made a mistake. And I thought it for good reason. Šêtu is a middle frequency verb meaning, “to remain, to be left over, to escape.” It works great in context. Šêṭu, on the other hand, is a less common verb whose meaning is in the general range of “to miss a target, to bypass, to disregard.” While “bypass” more or less fits the context, the center of the semantic range of šêṭu actually somewhat misses the target.
So I decided to take Alan’s wise advice and check my sources. Oppenheim and Butler both read ṭa. Just for good measure, I decided to see what CAD had under šêtu and šêṭu. I must have done this some time ago but I couldn’t remember what it said. The CAD šêṭu entry specifically references and quotes from the prayer I’m working on. It gives the same reading as Oppenheim and Butler. Well, believe it our not, at this point I still thought I might be mistaken. First, I’m sure Alan knows more Akkadian than I do. Second, it sometimes happens that scholars inadvertently perpetuate erroneous readings. Happily, Oppenheim provides a fairly readable photography of the only tablet fragment, K. 8583, containing any part of this word. K. 8583 is also the most complete fragment of the prayer. And lo and behold, the worrisome sign is ṬA (DA if you prefer) which, in theory, can represent either ṭa or tá (tá stands for a second, less common, alternative way of writing /ta/). So, if it were possible that here one can read the sign as tá, Alan might still be right. Two things make this very unlikely. First, Old Akkadian and peripheral Middle Assyrian/Babylonian are the primary witnesses for the ṬA sign equaling tá. The dialect of this prayer is more recent than any of these. Second, while only represented once in unbroken and indisputable context, the prayer uses the TA sign, and not ṬA, for /ta/. So, as I said, I thought I had made a mistake but I was wrong.
By the way, most of Alan’s other observations did point to real problems in my draft but none vexed me more than this one did.