Minor Sign Variants vs. Alternative Spellings

This page gives examples to distinguish these two concepts.

English "lite" and "light" are partly alternative spellings of the same "reading" and partly distinctive, having grown into distinct words with different ranges of readings. But even before they became different words, they *were* different spellings, and should be encoded differently in text data stored on computer.

The same can be exemplified for Cuneiform, in the first list below.

On the other hand, English "fi" and the "fi-ligature" are *not* distinct spellings, they are merely distinct typography. There are examples also of this in Cuneiform. Additionally, there are examples of single signs which differ so little that they can be considered typographic variants rather than distinct signs (characters of the script.

There is of course historical change in this, so that DIM x KUR and DIM x SHE were distinct signs in earlier script, but became mere typographic variants in later script. The infixed parts of these contain three vs. four tiny diagonal wedge-heads, so they differ in only one tiny mark. It is not surprising that they could lose distinctiveness. Other signs were not originally distinct, but developed by changing one of the wedges, or adding or dropping one, and so forth, and the difference never became distinctive.

Alternative Spellings with the Same Reading

There are a number of signs in two groups which look to be alternative spellings of the same words, with varying phonetic complements or semantic complements infixed or placed following the base Container sign, (Container signs can themselves function as a determinatives). Those comments are meant only to be suggestive, not a definitive analysis in these cases.

With the container NINDA2 there are the following sets of signs having the same readings but with distinct infixed components, or differing in whether those components are infixed or occur as separate signs following the base.
Those for which readings are listed as variants of $AM2 are the following set. Because they have the same reading, they may vary only in explicitness. If we were to encode the full sequence as a single sign (whether the elements are infixed or not), then those pairs of these which have all of the same components in the same order could be considered variants of a single sign. But if new readings are later discovered, that could refute this kind of analysis. The safe route is to code each of these as distinct. Higher-level software can then search them as equivalents, as desired, and ideally can permit users to specify their own sets of equivalences handling instances far beyond these.
NINDA2 x SHE B333v3 Deimel/Labat L187v5
(NINDA2 x SHE).A Deimel/Labat L187v4
(NINDA2 x SHE).A.AN Deimel/Labat L187v2
(NINDA2 x (SHE.A)).AN Deimel/Labat L187v1
NINDA2 x (SHE.A.AN) B333 Deimel/Labat L187
NINDA2 x (SHE.AN) B333v2 Deimel/Labat L187v3
NINDA2 x (A.AN) Deimel/Labat L190an
NINDA2 x AN B320 Deimel/Labat L178aa
NINDA2 x (SHE.A) listed only as an ELEMENT of $AM, not independently.

LAL (L481) and LAL2 (L482) both occur with infixes U, NI, or both. These are considered equivalents of each other. Yet the sets of infixed components are clearly distinct. So it would be relatively obvious to consider them distinct signs having the same reading:
LAL x U
LAL x NI
LAL x (U.NI)
LAL2 x U
LAL2 x (U.NI)

(KA x ME).GI citations from where?
KA x (ME.GI)

MUN$UB B820 L543 and MUN$UB2 B823 extended Deimel/Labat L543v2 differ by an extra component present in the second. So distinct spellings, warranting a separate encoding?

UB B504 L306 and $E$LAM B100 L065 are considered "graphic variants", but in this case it may mean that they may occur instead of each other in texts. Sort of like "liberty" and "freedom".

LU2 x KAM is given as the reading of two signs differing radically in the appearance of their infixed parts, in one case HI.BAD and in the other case KAM2 TENU (cf. B254 KAM2. -- B525. Should these be simply different signs used as mutual substitutes?

Minor Sign Variants

There are of course enormous numbers of these. Here only a couple of examples.

MUNU6 -- This appears to have SHE as its first component, but since the remainder of the sign does not occur independently, it is a single sign. The remainder of the sign after the "SHE" lookalike has several minor variants. See Ellermeier / Studt

GURUN B503 vs. B503v, L310 vs. 311.

The ligatures I+NA (Ellermeier / Studt page 41, Borger numbers B252lig and B252lig2) are equivalent and differ only in placement and orientation of three wedges out of nine, a different placement which has multiple analogs in variants of other signs. Since these are ligatures of the same base signs, they cannot have different character status, can only be variants (if one had status as a distinct character, it would no longer be a ligature, it would only have *arisen* from a ligature, as did the single letter which we can call <ae>, in Danish usage.

Component Differences Visually Minor

There are numerous examples in which one component of a sign of the type Container-x-Infix varies with a highly similar component, but the Container components are completely distinct in other cases, or when used independently without infixed components. These are one of the arguments for encoding Container-x-Infix signs as single characters, since the range of variation can be predicted only in the full context. It is the sign as a whole which has the variants, not the individual components of the sign.

BU CROSSING BU (BU crossed) has a different form in a complex sign than it does when standing on its own.

UTUA2 x GIR2 (= DAG.KISIM5 x GIR2) has a variant of the DAG part, where the same variation of the Container component is not registered in other signs. Extended Deimel/Labat L281av, L229an

SAR and EZEN are normally distinct, but SAR x BAD varied with EZEN x BAD.

Signs which perhaps through visual confusion have been used for each other may include the following. Such additional uses should have no implications for status as distinct characters of the components of the signs independently. If the sets of readings of the two characters do not overlap completely, then they retain a partial distinction, as distinct characters.

Examples with readings the same (not listed for the second sign form)
URU x TU varies with KA x TU Ellermeier p.18 -- B072 L039 vs. B027 L015xx
URU x MASH/BAR varies with KA x MASH/BAR -- B073 L040 vs. B030 L017an

Examples with readings distinct, or partly the same but not fully the same
URU x GU varies with KA x GU -- B084 L046 vs. B069 L034
URU x IGI varies with GI$GAL x IGI -- B079 L044 vs. B079v L044n

Resemblances between UM, DUB, URUDU may provide further examples relevant here.

B042, B042v GA2 …

Major Variants, Visually Highly Distinct

And now for some examples which gradually become more difficult, which are in some way intermediate between the other two.

An abbreviated form occurs of LU2 (Borger B514v, extended Deimel/Labat L330n, Ellermeier p.70) and of SHESH (Borger B535v, extended Deimel/Labat L331n, Ellermeier p.73). Probably simply to be regarded as a variant of the same character.

SIG7 = IGI energetic (called "gunu") arose from IGI with many parallel strokes above it (before the 90-degree rotation) or to the left of it (after that rotation). It has given rise to very distinct forms, still at least mostly used as equivalents. B564 vs. B654v, extended Deimel / Labat numbering L351 vs. L351n. This same variation occurs in at least some infixes in signs of the type Container-x-Infixed.

GURUN B503 and B503v, L310 vs. 311. These two are normal cuneiform signs. They differ by one wedge and are treated as variants of each other.
GURUN is also a reading of other sign forms. These are of the type "squared", four copies of the same component arranged around the sides of a square, in four 90-degree rotations. These may be historically related to B503, but are so different in form that they almost certainly have acquired distinctive status, and perhaps from each other. BAD squared (perhaps originally from IDIM squared?), and TAB squared -- B113a, B113b, L069x, L069xn.