Analyses Which Work for Limited Periods

There are quite a number of analyses of signs in various sources which work only for limited subperiods of Cuneiform script. These are manifest sometimes in naming of signs, sometimes in choosing what to count as an independent sign.

Given our principles for a standard cross-time-period list of signs, to keep all earlier distinctions which were later merged, and to keep all later distinctions which result from a single earlier sign, we will add single signs as needed. The following discussion is to make sure we are aware of these needs now, from the beginning.

"Gudea" below is not always the correct reference, when classical cuneiform is intended.

Sequences in Neo-Assyrian which Were Single Signs Earlier

UMBIN "claw, talon" in Neo-Assyrian looks like GAD.KID2.UR2. But that analysis does not work for old or in middle Babylonian. Since analysis as a single sign will be necessary in any case for the earlier script, it should be encoded now. In Classical Sumerian the source is clearly revealed as some form of ("leg" x KID2), where KID2 is a kind of hand or grasp, a single sign which can be regarded as SHU TENU. Another Classical Sumerian version seems to have reanalyzed the upper part as GAD, so it looks like GAD.(UR2 x KID2). -- B160 L92b
Ellermeier’s catalog also lists a form UMBIN as (GAD.UR2), which could support an analysis (GAD.UR2) x KID2, and a form B158 L82 GAD.KID2. More information about the status of these two might be relevant. Other sign names used here: AKKIL, SIG8.

LU2 "$e$$ig" is not that, but rather (LU2 with falling marks). This is shown by its later equivalent LU2.GAM or RAB.GAM, where GAM has meanings in the range "kneel, bend, fall, die". -- B516 L330 and B518
The early (LU2 with falling marks) x BAD has the later equivalent (LU2.GAM) x BAD or simpler (RAB.GAM) x BAD. Since the early single sign (LU2 with falling marks) is needed in any case, it is the better encoding for the later (LU2.GAM) or (RAB.GAM).
The analysis (LU2 x BAD).GAM is not etymologically valid, and needlessly disrupts the relation of older and later stages of the script.

IDIGNA was originally not a compound if correctly identified in ZATU -- B124 L74x

"SHE.NAGA" was a single sign in Gudea, Fara, and Uruk, B293v L375a ZATU381.

SAG energetic (not SAG gunu) is not RA.PA, it is a single sign. -- L329 DUL3

KAD4 was not SHU.KAD2, as seen in Gudea -- B568 L354b
KAD5 was not SHU.KAD3, as seen in Gudea -- B569 L354bv

IL2 is not a sequence of parts (GA.DUN3) as shown by Classic Sumerian and Uruk, where its lower part is seen to be LU2 -- B493, L320

"MA2.KASKAL" is disproved by Gudea and ZATU -- B203 L122b

"AM.GISH" is not a valid analysis of NE

Signs Whose Infixed Components are Distinguished Only Earlier

Ellermeier reports that earlier NINDA2 x MASH should now be called NINDA2 x BAN2, presumably because it reflects the earlier form which is unambiguous, where the later form cannot distinguish MASH from BAN2.

The multiple merger into KU can be resoled in various ways.
KA x KU is revealed by Gudea typography as KA x SHE3 -- B062 L33

NIN is (SAL x TUG2) later, (or x KU, merged), but SAL x SHE3 or SAL x NAM2 in Uruk -- B897

SHE3 (think of a strip of woven cloth) can be confused with GAN2 or angled KAR2. Where TUG2 (cloth) varies with KAR2, we may possibly be able to assume an early SHE3 as their origin.

GAN2 and KAR2 are merged into GAN2 when infixed. Can they sometimes be distinguished? Or was the difference mostly neutralized even early, except as independent signs.

ZI2 is a late fusion, revealed as earlier AB x PA. -- B259 L147 single sign

A2 = DA x SHE in Gudea (conceivably DA x NUN earlier?) B560 L334

DIM x SHE and DIM x KUR are distinct in Early Dynastic usage -- B168 L95 ZATU368 LAK55 vs. LAK56

LAGAB x UD and LAGAB x ERIM both have Borger’s number B783, but different names, presumably because there is early evidence of a distinction.

ASH+ASH This ligature looks later like HAL. These are distinguished by Borger as infix to NINDA2.

LAM x (KUR.RU) is revealed by Gudea

Apparent Container-x-Infix Signs Whose Atomic Nature is Revealed Earlier

UKKIN is not URU x MASH, as might be thought later. It is an atomic primitive. -- B073 L40

BIR is not (HI x GAN) -- B643 L400 Fara 372
HAR is not (HI x ASH2) -- B644 L401 Fara370

GUG2 is not (SHE3-gunu, as seen in Gudea, Fara790, and probably ZATU237 = LU3 ?

Kerning? Or status as a single sign more evident in earlier times: IGI, SAL, RU

These can be analyzed as single signs, even though their ultimate etymology may be recognizable, or they can be analyzed as sequences with obligatory ligaturization. What is the empirical evidence on freedom of dividing these across line breaks? Or could this be a matter of kerning, for signs like

ABZU in Gudea typography is either ligatured or fused. Fara 292 Combination occurs already in Uruk, but not ligatured -- ZATU17

IGI and SAL which have a point to the right which can contact a next sign without , or RU, which has a valley which can surround the point of a sign to its left. More on this type later.
IGI x RI is more evidently a single sign AR in Gudea -- B726 L451 Fara422
IGI x RU is more evidently a single sign PAD3 in Gudea -- B725 L450 Fara423 and in Uruk

Splits late

KALAM B500 vs. UN B501 had the same form originally?
TA vs. TA*
DUH / DU8 L298v is in NeoBabylonian distinct from GAB L298 v2.