In two common Sumerian stock phrases, there is the problem of the word níŋ-U.GUD, which can be read as either níŋ-ul or níŋ-du7. The first reading means something like “ancient/traditional things”, and the second reading means something along the lines of “propriety” or “appropriate things”.

The two common formulae are:

(1) níŋ-U.GUD-e pa        mu-na-è
    niŋ??=e     pa=0      mu-na-e-0
    ??=LOCT     CVNE=ABS  CP-DAT3-manifest-ABS3I
    'For him, he made manifest the ancient things.'
    'For him, he made manifest the appropriate things.'


(2) ur-saŋ   níŋ-U.GUD-e gù        ba-a-dé
    ursaŋ=0  niŋ??=e     gu=0      ba-?-de-0
    hero=VOC ??=LOCT     voice=ABS CP-ERG2-pour-ABS3I
    'Hero, you called for (=pour voice) for the ancient things.'
    'Hero, you called for propriety.'

Formula 2 appears in the Stele of the Vultures of Eannatum (albeit in slightly altered form):

é-an-na-túm á tuku-e mu-pàd-da <d>nin-ŋír-sú-ka-ke4
'I, powerful Eannatum, called by Ningirsu,'
é-an-na-túm-me kur a-ne-šè ŋá-̵ŋá-dè
'(I) Eannatum, in order to lay down the land before him,'
níŋ-ul-lí-a-da gù nam-mi-dé
'I called for (?/ proclaimed) the ancient things!'

The main difference beyond the use of the comitative instead of the locative-terminative is that in this instance, we get a longer form of ul (ul-lí), wherefore we know that du7 is not a plausible reading for this instance.

Formula 1 gets a similar break in Ur-Namma:

<d>nanna dumu-sag <d>en-líl-lá lugal-a-ni
'For his master Nanna, sweet son of Enlil:'
ur-<d>namma nita kal-ga lugal úri<ki>-ma
'Ur-Namma, a mighty man, King of Ur,'
lugal ki-en-gi ki-uri-ke4 lú é <d>nanna in-dù-a
'King of Sumer and Akkad, who built Nanna's Temple:'
níŋ úl-lí-a-ke4 pa mu-na-è
'He made manifest the things of the distant past.'

Again, there is a slight difference from the more common formula (the extra genitive, uli-ak=e instead of uli=e). But the basic meaning here is quite clear.

Because we find both Formulae (with slight variation) occasionally using longer stems of ul, which are incompatible with a reading du7, it is tempting to say that the correct reading is ul instead of du7 in all instances of these formulae. But I err on the side of caution: in all instances which have only níŋ-U.GUD-e, du7 as a reading for U.GUD works just fine semantically. So, either ul is the correct reading in all cases, or there are variants of the formulae which also refer to propriety.