Conventions for naming signs in terms of their components

The distinction between "sign" and "component" is intended in the same sense as used for Han characters.
Parentheses are here used to group names of components in order to reveal the structure of signs. If there are no signs whose names differ only in the differences signaled by the parentheses, then the parentheses can simply be dropped, since they are not permitted in Unicode names. The syntax of names is however something we should be consciously aware of in the preparation of a sign list, since structural questions do signal to us other problems in encoding.

One of the major problems in sign names is that if first level signs are decomposed into components, as TUR3 into NUN x LAGAR, then the paired form is much more difficult to express and does benefit greatly from (might require) parentheses. The difference is
TUR3 OVER TUR3 (or TUR3 PAIRED) vs.
(NUN x LAGAR) OVER (NUN x LAGAR)
(or NUN x LAGAR PAIRED, if we can use a convention that the term "PAIRED" always applies to the entirety of what precedes)

SIGN variant
A sign form which represents the same sign as another form, which is listed only for convenience in cataloguing. To be considered a variant, it is not sufficient that the reading be the same as that of another sign form. The variant must also truly be a variant only in glyphic form, with a close relation to the primary form of the sign of which it is claimed to be a variant. If a given functional reading can be spelled in two different ways (with recognizably different components which are otherwise known to be distinct signs), then the form in question is a distinct sign, not a mere variant. This is much like English "lite" vs. "light". Even if there were no distinction at all in usage between these two, they would still be encoded differently.

SIGN + SIGN refers to true ligatures, which are equivalent in readings and in meaning to a sequence of separate signs not ligatured. These are analogous to Latin ligatures "fi, ffi, fl, ffl". Ligaturing is not to be used as a device to artificially decompose single independent signs into parts which are merely similar or the same in appearance..
SIGN x NUMBER combinations may for the most part usefully be considered ligatures instead, SIGN + NUMBER, since they have the same meaning as their individual signs. There are two kinds of exceptions to this: Strokes marking the ages of young animals are not numbers counting the animals. These are counted as separate signs, much as English "yearling" is a distinct lexical item, not a combination involving the number "one". The sign for 'day' U4 has three different positions for associated numerals, depending on whether counting days, months, or years. So the positions of the numerals must be distinguished in some fashion. (Can all three numeral notations occur with a single instance of U4?)

SIGN x SIGN refers to single signs whose components are normally a "Container" and one or more "Infixed" parts. Just as for Han characters, "container" components need not fully surround their "Infixed" components. The components have the same names as the independent signs to which they are related, and (at least normally) all components should exist as independent signs. In older cuneiform, the "infixed" component(s) are often affixed to some part of the "container" component, or overlap it, rather than being strictly inside it.

SIGN CROSSING (substituting for all instances of SIGN CROSSING SIGN, because the signs are identical in all of the cases in which this naming pattern was used, there is no independent choice of the second sign)

SIGN PAIRED (substituting for all instances of "SIGN OVER SIGN", because the sign component is almost always the same. A single case is known of SIGN OVER SIGN with the two signs distinct, namely KU OVER (HI x ASH2), and for that one the name is retained). This has the advantage that the name works both before and after the 90-degree rotation in the script.

SIGN SQUARED (four copies of a sign, arranged around the sides of a square)

SIGN with SIGN
SIGN with SIGN above
This naming pattern is substituted for some instances which have been called SIGN x SIGN where it seems more appropriate. The phrasing "SIGN with SIGN above" works only before the 90-degree rotation of the script -- later it refers to what we would call "SIGN with SIGN to the left".

The number of distinct infix components or complexes are given in parenthesis just after the name of the "Container" component. The total number of signs here classed as of the structure (container component x infixed component(s)) is 905, a preliminary count I will modify in a couple of days.

Containers which completely surround infixed component(s) in later or earlier cuneiform or usually both if the signs occur throughout all time periods (surrounding on both sides and either top or bottom is counted as surrounding on all sides):
AB (17)
AG (3) '
AL (7) 'hoe, pickaxe'
DA (1) 'arm, side'
DIM (1)
E2 (8) 'gate'
EZEN (43) 'festival'
GA2 (102)
GESHTIN (2) 'wine, grape'
GISH (1) {determiner for trees, wooden objects}
IL (1) 'tall' {= EGIR sheshig}
IR3 (1) 'male'
ITI (1)
KA (68) 'mouth' etc.
KAK (2+) 'build'
KI (7) 'land, domain'
LAGAB (83)
LAGAR (8) 'priest'
MA (2) 'fig tree'
MUN (1) 'alt, potash'
NE (2) 'fire'
NIM (3)
PISAN2 (3) 'jar'
SHA3 (10) 'heart'
SHAR2 (5) '3600' – infixed components are numbers 10, 20, 30, 40, 50
SHID (2) 'accountant'
SHA (2)
SHITA (6)
TUR3 (2) 'household' 'livestock corral'
UR (3) 'dog' in later cuneiform
UR2 (7) 'haunch'
USH (4) 'male'
ZATU85 (3) {DU8 ?}
ZATU366 (1)
ZATU651 (8)
ZATU714 (1)
ZATU737 (15)
ZATU762 (2)
The following three are used as container components almost only in archaic.
DUG (48) {one form DUG x KUR fused to yield a later sign}
GAN (13)
ZATU341 (13)

Total in this category: 512 signs

Containers which surround infixed component(s) on left, top, and bottom in later cuneiform (one example only on top and bottom), but in earlier cuneiform either surround infixed components on all sides, or else the "infixed" component occurs as an affix to some part of the "container" component. NINDA2 surrounded three sides of infixed component(s) also in oldest cuneiform.
AB2 (15) 'cow'
BI (3) 'beer' {= KASH}
GUD (2) 'ox'
HUBUR (13) (originally some eared animal)
LU2 (23) 'person'
LU2 SHESHIG (1)
NINDA2 (48)
PIRIG / GIR3 (~13) 'lion' and other animals
UTUA2 (27) 'bull' and 'insect'

Total in this category: 145 signs

Containers which surround infixed component(s) in earlier cuneiform, but in later cuniform have extruded the infixed component(s) to stand outside the container component (and one example where reanalyzed as parts which no longer represent the sign origins)
GA (5) 'milk'
GU2 (11)
HI (15) {later merges with SHAR2}
LAL (2 ?) 'lack'
LAL2 (9) 'bandage'
TAG (9) 'cut throat'
UDU with TEATS '(female) goat' vs. later analysis MA2.KASKAL

Total in this category: 51 signs

Containers which in later cuneiform surround infixed component(s) on left and right (so the container component appears split). Some of these in older cuneiform surround infixed component(s) completely.
A (13) 'water' {split around infixed component also in older cuneiform} For a list please click here.
DUB (5) 'tablet, record' For a list please click here.
HUB2 (5) 'dance'
LAM (2)
MUSH3 (5) 'Inanna'
SAG (20) 'head'
SHEN (1)
SHIM (14) (determinative for) 'aromatic plants'
TA (3)
UD (2) 'day; white'{UD x MI decomposed in later cuneiform ???}
UKKIN (~11)
UM (4) For a list please click here.
URU (34) 'city'

Total in this category: 119 signs

Containers which appear to consist of two components, but are treated as a single sign and permit infixation of other components between the parts
(HU WITH UD) x SHE
NAGA (1) {looks like SUM, a sign occurs with SHU TENU45 infixed}

Containers which occur to the left and top of an infixed component, or to bottom and right of an infixed component, the infixed component "nested" in the niche of the container component,
EN (10) {a chief administrator}
PI (10) 'ear' or 'basket'?

Total in this category: 20 signs

Containers which occur to the left of an infixed component (which before the 90-degree rotation of the script occurred above it) Compare HAN characters radical 'roof'

SHU2 'cover, roof, vault of sky' For SHU2, Labat numbers at least 546-548, 550, 553. For details, please click here.

SHU4 'cover, roof, vault of sky' (same, over narrower base sign) For SHU4, Labat numbers at least 413 through 419, 428-430, 441-443, 447-448, 469. For details, please click here.

Total in this category (so far): 21 signs

Containers which occur below the infixed component in older cuneiform (and would be to the right of it if they survive(d?) into later cuneiform)
SILA3 (37)
RU may be a "container" of this type; or it may be merely ligatured with the elements which it has nested inside it, in cases of different signs. not yet counted.

Total in this category: 37 signs

Combinations of Container component with Infixed component which fused in later cuneiform, so components no longer visible
DA x SHE 'wing, arm, side; right; force'
GUD x KUR fused in OB, OA
LAGAB x PA 'take possession'
RA 'bathe, immerse' is composed of something with original lower part SI (*tank of water?)

Two examples of components which occurred in cuneiform before the 90-degree rotation quite often above a base, sometimes adjacent to it, and are to be analyzed as modifiers, not as the primary component of the sign, so included here as "with SIGN"
"with GAL (above)" as in LUGAL fused from older LU2 with GAL above
"with SHE attached" as in NAMSHE from NAM2 with SHE attached above ("above" before the 90-degree rotation)