ARB š- poses a problem: it can be from *s- influenced by -l; otherwise *ŝ1 {} *ŝ has to be reconstructed in the proto-form.

    Note MND šalhapta d_-mahria [DM 442] translated 'Ransom of Disease'.

    Cf. [Maizel 176].

    Cf. šālāl šäl bēṣīm 'embryonic eggs laced together with veins; ovary of birds' [Ja 1585]

    Both ETH s- and MHR h- (quoted only by Leslau) from š, together with SOQ š-, imply *š- {} *s-; another MHR example (also attested in a strange way) and both JIB examples point to *ŝ- {} *ĉ-.

    A plausible semantic connection and a rare combination of radicals speak in favor of a relationship, rather than a chance coincidence with HBR mošzār 'twisted' (hof, part.) [KB 959], ARB šzr 'retourner la main à droite en brandissant la lance; tordre, donner un tour de droite à gauche à la corde' [BK 1 1225] and, with a further meaning shift, MND šanziria (pl.) 'the ropes' [DM 443].

    It is difficult to avoid comparison of this root (as a form with suffixed -r?) to AKK šizû 'one-third cubit' OB on [CAD š3 152], [AHw 1254] (also šizum, šizât- [ibid.]); the latter, however, may be alternatively compared to ARB šd_d_ 'être seul, isolé, séparé des autres; séparer, isoler' [BK 1 1206] and GEZ ŝazaya et al. 'to detach, separate' [LGz 541] <*ŝd_d_/y {} *ĉǯǯ/y (cf. also SOQ šéze 'se séparer, etc.' [LS 55], which Leslau considers causative-reflexive of ʔézi [ibid.] comparing at the same time to GEZ ŝazaya; note that -z is from *z, and not *d_).

    Cf. what looks like a variant root in TGR sädǝr '(hand-)span', säddärä 'to measure by spans' [LH 196] and TNA säddärä do. [ibid.] to be compared to MND ŠDR 'to twist' [DM 450], šidra 'twisting, spinning by hand' [ibid. 460].

    Note that the present root and ARB šibr- compared by Pretorius (see discussion in [LGz 523]) are not related. ARB šibr- 'empan, mesure de longueur que donne la distance entre le pouce et l'articulaire étendus' [BK 1 1183] is to be compared to MSA: MHR ŝēbǝr 'span of outspread fingers' [JM 372], HRS ŝéber [JH 118], JIB ŝɛ̄r, pl. ɛŝbɔ́r [JJ 245] do. The ARB and MSA examples are not made into a separate entry because the latter one can be borrowing of the former, in view of no cognates in other SEM

    Note šippūlē mēʕayim 'lower part of the abdomen' [ibid.]

    Corresponds to HBR yark_ā in the meaning 'backside' [KB 405]

    Cf. mišfalat- 'gésier, jabot (d'oiseau); estomac' [BK 1 1250], with š instead of the expected s (š- <*ŝ-, "lateralization" <*s- influenced by -l?)

    Cf. šefle 'having large feet' [LH 231]

    Cf. different forms in ARM derived from this root or directly from *špl 'to be low': JUD šippūlā 'extremety, bottom, train' [Ja 1556] and SYR šǝpōlā 'pes (montis); fundamenta' [Brock 795].

    Note an unusual five radical noun pattern šǝnfǝlla in ETH, with a secondary -n and doubling of the last radical not typical of SEM (except AKK).

    Note MOD ETH (TGR, TNA, AMH, END) *sambar- 'stomach, part of the stomach' [LGur 546] (a variant root?).

    [Holma 161-2]: AKK, ARB (sāfilat-, safilat-)

    Note that the word is used in particular in the dual išdān which is found in AKK only with nouns denoting body parts or their semantic derivates

    Cf. sath-, satah-, sitah- 'derrière, cul' [ibid. 1051] (on h as a triconsonantizer, see Introduction)

    According to Leslau, from CUSH (Hadiya suto), but hardly so in view of the comparative data

    The final -ʔ in the two latter forms is difficult to explain. Cf. also [SSL LS 1475; 1476]

    Note that in JIB and SOQ, the last radical -t is perceived as a feminine marker and dropped (cf., however, its reappearance in dual and pl. in SOQ).

    Presumably a form with the *-t suffix < *ʔVšš- {} *ʔVss- 'foundation, base' (with -šš > -š before -t?): AKK uššu 'Fundament' [AHw 1442] (according to [AHw 1442] and [Kauf 110], borrowed from SUMER uš₈ 'Fundament' [IK 1149] and then to ARM to ARB; cf. also e(s)sû 'Niederung' SB [AHw 250]); HBR PB ʔūš 'foundation' [Ja 35] (<JUD?); ARM: BIB ʔuššayyā, pl. (du.?) det., 'Fundament' [KB deutsch 1675], JUD ʔuššā, ʔawšā 'foundations, fortification' [Ja 35]; ARB ʔiss-, ʔuss-, ʔasās- 'base, fondement' [BK 1 30-31]; SAB ʔs₁ 'base' [SD 7]; TGR (?) ʔässäsä 'to set in order' [LH 363] (where it is compared with ARB ʔassasa 'to lay the foundation'); JIB sɔs 'foundation' [JJ 220] (according to Johnstone < *ʔss; OMAN ARB sās quoted [ibid.] admits a loan from ARB).

    Note that in JIB and SOQ, the last radical -t is perceived as a feminine marker and dropped (cf., however, its reappearance in dual and pl. in SOQ).

    As to the AKK, EBL and UGR forms with -d, one wonders whether these should be made into a separate root *ʔišd- {} *ʔisd- 'foundation, bottom, lower extremities' or explained from *ʔiš(V)t- by contamination with *(ʔi-/wi-)sād(-at)- {} *(ʔ/yV-)cad(-at)- 'base, foundation': UGR msdt 'Grundfeste' [Aist 131] (placed under *ysd); HBR yǝsōd 'foundation wall, base' [KB 317], ysd 'to found, establish' [KB 417]; JUD ʔīsādā 'head-side, pillow' [Ja 53], SYR ʔesādā 'a capite, sub capite' [Brock 32]; ARB ʔisādat- 'coussin' [BK 1 31], wisād(at)-, wasād(at)- 'coussin, oreille; lit de repos' [ibid. 2 1533] (cf. also sady- 'trame d'une tissue' [ibid. 1 1074]); SAB ms₃d 'base, plinth of statue' [SD 163] (placed under *ws₃d); TGR cf. sudot 'nuque' [LH 197, apud d'Abbadie] (primary meaning 'base of the head'?); MHR sáddǝt 'raised platform, island in a valley' [JM 341], JIB (EAST) sédt do. [ibid.].

    [Fron 46] (*šit- 'natica' /JIB,ARB,SYR,HBR,UGR,AKK/); [Holma 128]: AKK, HBR, SYR, ARB; [DLU 56]: UGR, AKK, EBL; [KB deutsch 1536]: HBR, UGR (št), AKK (išdu), SYR (ʔeštā), ARB (ʔist-, wisād-), PHO (rejecting connection with ARM ʔuššā and AKK uššu); [LS 413]: SOQ, MSA, ARB, HBR, AKK

    Found in one LL where it is equated to ubānum. Considered a W. SEM loan without definite source [CAD] or a derivate from ṣabātu 'to catch' [AHw]. Note that if a loanword, ne- prefixation is difficult to explain, while if a genuine AKK word, it can be regarded as a prefix *ma- (on which see Introduction) which regularly shifts to na- (> ne- because of *-ʕ) if one of the root consonants is a labial

    Also ṣbʕ 'certain measure of length, finger' [ibid. 958]

    Cf. also ʔädbaʕ (ʔädbǝʕā) 'finger' [Ja 15], with an unexplained -d- (a variant root?)

    Forms without ʔV- are attested in many ARB dialects

    This form quoted [ibid. 108] and in [LGz 45] is questionable both phonetically and semantically (-ṭ < *-t influenced by ṣ-?); the form ṣatä quoted in [LGaf 237] looks phonetically more tenable

    -ḥ < -ʕ in Auslaut. Cf. also [SSL LS 1451]

    Note a meaning shift in ESA: SAB ʔṣbʕ, used to denote fractions, e.g. ʔṣbʕm bn t_mny '1/8'' [SD 140]; MIN ṣbʕ 'doigt' ("exprime les termes d'une unité") [LM 93]. [DLU 55]: UGR, EBL, ARB, GEZ, HBR, ARM, AKK; [KB 81]: HBR, UGR, ARM, ARB, SAB, GEZ, AKK (iṣbittu, not found in AKK dictionaries); [Brock 620]: SYR, HBR, JUD, ARB, GEZ; [LGz 45]: GEZ, ETH, ARB, ARM, HBR, UGR, ESA, SOQ; [LS 70]: SOQ, MSA, ARB, HBR, SYR, GEZ

    Goes back to *ṣarraɣ-t-, which, in its term, may originate from *ṣVrnaɣ-t- or to be formed after the ḳattal-t pattern widely used in HBR to denote deseases (cf. gabbaḥat 'baldness', dalläḳät 'burning fever' etc.); in this case, however, shaping of an earlier quadriradical base after this widespread pattern is also plausible

    < *ṣinɣ-at- <*ṣinnaɣ-at- <*ṣirnaɣ-at-?

    Note *ʕ > ʔ in the second example

    *-ɣ is hypothetic, as in all the languages listed ʕ may yield SEM *ɣ or *ʕ, while the ARB example with -ɣ is not reliable because of -m- instead of the expected *-n-.

    [Fron 40] (*ṣarraʕ-at- 'malattia della pelle' /GEZ,JUD,HBR,AKK/); [KB 1057]: HBR, JUD, AKK, GEZ; [LGz 564]: GEZ, HBR, AKK

    Hapax, interpretation uncertain (not in [CAD]); in [Gelb OAkk 304] quoted as zawār- and left untranslated

    Cf. ARB ṣwr 'tourner le visage d'un côté' [BK 1 1383] (from 'to turn the neck'?), likely related.

    Note MHR ṣ̌ārī́r 'nape' [SSL 1 291]; however, since MHR ṣ̌ usually (though not always, cf. ṣ̌ǝbáʔ 'finger, toe' < SEM *ṣibaʕ-, and ṣ̌ǝfdēt 'frog' vs. ARB ḍifdiʕ-) corresponds to SEM *ḳ, it may be more safe to compare the MHR word with SOQ ḳar 'cou' [LS 384] (cf. also MHR ḳār do. quoted [ibid.] apud Jahn, but not found in [JM]).

    One wonders whether this is the source of ETH *ṣwr {} *c̣wr 'to carry (on shoulders ?)': GEZ ṣora 'to bear, carry, support, etc.' [LGz 567] and other ETH [ibid.]; note that in GUR this verb means 'to carry loads on the back or on the head or on the shoulders' [LGur 637].

    [KB 1009]: HBR, ARM, AKK, GEZ (ṣwr 'to bear'), ARB (ṣawr-); [Brock 625]: SYR, HBR

    Note a stable -ā- instead of the expected -ē- (<*-Vʕ-). Cf. šuʔru, šūru 'eyebrow' OB, SB [CAD š3 366], [AHw 1287]

    Possibly an old apophonic plural *ŝiʕar- from sing. ŝiʕr-; may it be reflected in the nomen unitatis ŝaʕărā [ibid. 1345] < *ŝaʕrat- < *ŝiʕrat- (*-i- >-a- before ʕ) ?

    -ǝ- of the first syllable is probably from *-a- before ʕ followed by ǝ

    Scarce evidence in Southern SEM area only.

    Note ARB šabḥ- 'large' [BK 1 1182]; a semantic shift 'fat' (n.) > 'fat' (adj.) > 'big, large' is not to be excluded. Cf. also ARB ḥāšib- 'qui a le ventre large' [ibid. 431], likely related with metathesis.

    Cf. *ŝaḥm- {} *ĉaḥm- (No. ).

    [LGz 525]: GEZ, ETH, JIB, HBR PB šbḥ 'grow in value', SYR šbḥ 'improve, raise in value' (in both latter examples, š is <*š {} *s and uncomparable either phonetically or semantically)

    Both translated as 'ein Art Fleisch od Wucherung ?' in [AHw 1211]

    Note the comments: "It is cut in a slaughtered animal to see how much fat is on the carcass" [ibid.]

    Cf. *ŝabḥ- {} *ĉabḥ- (No. )

    The irregular -ē-/-ī- (instead of the expected *-ā-) may be explained from the change *ŝimaʔl- > *šimiʔl- > šumīl- (*-i- > -u- near -m-)

    To be normalized as ši/umaʔl- ?

    Note -ā- in ŝǝmā(ʔ)lī 'left' (adj.) [ibid. 1333], which is usually explained as a regular reflex of *ā in an unstressed position (versus -ō- in ŝǝmō(ʔ)l where *ā is stressed). It seems more plausible, however, not to connect this discrepancy with accentual phenomena, but rather to regard the adjective as derived after a different pattern (probably *CaCaCiyy-, сf. yǝmānī 'right' with -a-, which neither can go back directly to its presumed prototype yāmīn 'right side')

    Orthographically also with both ŝ and (ʔ) ЛУЧШЕ ИХ ПРИВЕСТИ [Ja 1591], [Sok 571]

    Also asmala, ʕsmala (with the *ʔa- prefix)

    On the possibility of suffixed *-l see Introduction; cf. ARB šaʔmat- 'côté gauche' [BK 1 1179], SAB s₂ʔm 'North' [SD 130] and JIB ŝiñ (a correct transcription is ŝĩn, see [SSL 2 247]) 'gauche' quoted in [LS 64] under ʔímhel (influenced, according to Leslau, by iñ 'droite', cf. notes to *yamīn~ *yamān-, No. ).

    See metathesis with a meaning shift in ARB šaʔamal- 'vent du Nord' (also šamʔal- do.) [BK 1 1273] and ESA: SAB h-s₂ʔml 'be northward' [SD 130], MIN s₂ʔml-s₁ 'vers le nord' [LM 85] (note s₂mʔl-s₁ quoted [ibid. 86] as a contextual form).

    Note BERB *a-zǝlmaḍ 'left', *z- yielding AFRAS *ĉ. With three identical root consonants out of four, though in a metathetic order, and such specific meaning as 'left', the SEM and BERB roots are very likely related; BERB -ḍ in Auslaut, though, remains unexplained (<*a-zǝlmaʔ-t <*-ĉilmaʔ-t < metathetic *ĉimʔal-t ?).

    [Holma 3]: AKK, HBR, ARB, SYR; [KB 1332]: HBR, UGR, AKK, ARM, ARB, ESA; [Brock 481]: SYR, ARB, HBR, AKK; [LS 430]: SOQ, MSA, ARB, HBR, SYR, AKK

    Translated as 'tip of an animal's horn' in [CAD s 165] (note s- instead of the expected š-)

    Note, however, that another meaning of the same word is 'bord, extrémité, crête'; possibly a contamination of two roots one meaning 'edge' and the other 'eyelash'

    Cf. AMH šǝfal 'eyebrow', a variant root with l/r?