← Semitic Search

Sound correspondences

Empirical reflex table derived from 171observed phoneme-pairs across the 164 usable editor-curated cognate claims. Each row reads: “when the left language has this surface phoneme, the right language’s cognate tends to have that phoneme”. Identity correspondences make up 53% — the rest are the Proto-Semitic sound changes that matter for cross-script cognate discovery.

Proto-Semitic reflex matrix

How each Semitic language realizes the nine key Proto-Semitic consonants that split differently across branches. Dashed cell means the phoneme was lost or merged out of the inventory. Where two reflexes are listed, the split is conditional (position / environment). Sources: Lipiński (2001), Huehnergard (2000), Moscati (1964).

lang*ṯ*ḏ*ḍ*ẓ*ḫ*ġ*ś*š*ṣ
Akkadianakkšzṣṣḫ-ššṣ
Ugariticugṯḏṣẓḫġś/ššṣ
Arabicarṯḏḍẓḫġšsṣ
Hebrewhešzṣṣḥʿśšṣ
Syriacsyctdʿṭḥʿsšṣ
Imperial Aramaicarctdq/ʿṭḥʿsšṣ
Ge'ezgezszḍṣḫʿśsṣ
Sabaeansabṯḏḍẓḫġśs₁ṣ
Phoenicianphnšzṣṣḥʿsšṣ
preservedmerged with another phonemelost

Evidence density grid

How many phoneme observations connect each source → target language pair. Cells are colored by density — dark = more evidence. Blank cells mean no editor claims cover that direction.

src \ tgtakkamararcgezhesycti
akk—·3·····
am·—6·····
ar··—··329·
arc··8—·33·
gez··18·—···
he8·62··—12·
syc··10···—·
ti··6····—

Empirical reflex weights from editor claims

Akkadian → Arabic

3 observations
akk→arprobability
b→ḥ50%reflex
b→b50%
m→b100%reflex

Amharic → Arabic

6 observations
am→arprobability
m→w50%reflex
m→m50%
n→n100%
r→r100%
z→z100%
ʿ→ʿ100%

Imperial Aramaic → Arabic

8 observations
arc→arprobability
b→b100%
ḥ→ḥ67%
ḥ→ḫ33%reflex
l→l100%
m→m100%
n→n100%
š→s100%reflex
w→y100%reflex

Imperial Aramaic → Hebrew

3 observations
arc→heprobability
ḥ→ḥ100%
m→m100%
š→š100%

Imperial Aramaic → Syriac

3 observations
arc→sycprobability
ḥ→ḥ100%
m→m100%
š→š100%

Arabic → Hebrew

3 observations
ar→heprobability
n→n100%
ṯ→š100%reflex
y→h100%reflex

Arabic → Syriac

29 observations
ar→sycprobability
b→b100%
d→d100%
ḍ→ʿ100%reflex
f→b44%reflex
f→f38%
f→l13%reflex
f→r6%reflex
g→g100%
ḥ→ḥ100%
ḫ→ḥ100%reflex
l→l80%
l→ḥ20%reflex
m→m100%
n→n100%
q→k54%reflex
q→q46%
r→r100%
s→š100%reflex
š→f40%reflex
š→š40%
š→s20%reflex
ṣ→ṣ75%
ṣ→s25%reflex
ṭ→ṭ100%
w→w100%
y→w50%reflex
y→y50%
z→z100%
ʿ→ʿ100%

Ge'ez → Arabic

18 observations
gez→arprobability
b→b60%
b→q20%reflex
b→h20%reflex
f→f100%
g→w33%reflex
g→b33%reflex
g→g33%
ḥ→ḥ100%
l→l100%
m→m67%
m→w33%reflex
n→n100%
s→h67%reflex
s→s33%
t→ʾ100%reflex
y→y100%
z→z100%
ʾ→ʿ100%reflex

Hebrew → Akkadian

8 observations
he→akkprobability
b→b100%
f→p100%reflex
ḥ→h100%reflex
l→l100%
n→n100%
ṭ→t100%reflex
w→b100%reflex
ʾ→n100%reflex

Hebrew → Arabic

62 observations
he→arprobability
b→b84%
b→f11%reflex
b→m5%reflex
d→d92%
d→w8%reflex
f→f100%
g→g100%
h→h31%
h→y23%reflex
h→w15%reflex
h→b8%reflex
h→d8%reflex
h→ʾ8%reflex
h→r8%reflex
ḥ→ḥ73%
ḥ→ḫ27%reflex
k→k100%
l→l92%
l→ʾ8%reflex
m→m100%
n→n91%
n→š4%reflex
n→ḍ4%reflex
q→q93%
q→t7%reflex
r→r89%
r→w7%reflex
r→m4%reflex
s→z67%reflex
s→s33%
š→s69%reflex
š→ṯ15%reflex
š→y8%reflex
š→š8%
ṣ→ḍ50%reflex
ṣ→ṣ50%
t→t100%
ṭ→ṭ33%
ṭ→ẓ33%reflex
ṭ→t33%reflex
w→w50%
w→b25%reflex
w→ẓ13%reflex
w→f13%reflex
y→w56%reflex
y→y22%
y→h11%reflex
y→r11%reflex
z→z60%
z→ḏ20%reflex
z→ẓ20%reflex
ʾ→ʾ69%
ʾ→ḥ6%reflex
ʾ→ʿ6%reflex
ʾ→ġ6%reflex
ʾ→n6%reflex
ʾ→w6%reflex
ʿ→ʿ42%
ʿ→ġ25%reflex
ʿ→q17%reflex
ʿ→t8%reflex
ʿ→g8%reflex

Hebrew → Syriac

12 observations
he→sycprobability
d→d100%
f→f100%
g→g100%
ḥ→ḥ100%
m→m100%
q→q100%
r→r100%
ṣ→ʿ100%reflex
t→t100%
ṭ→ṭ100%
w→w100%
z→z100%

Syriac → Arabic

10 observations
syc→arprobability
b→b100%
d→d67%
d→ḍ33%reflex
g→g100%
h→h100%
m→m100%
n→n100%
ṣ→ṣ100%
y→w100%reflex
ʿ→ʿ100%

Tigrinya → Arabic

6 observations
ti→arprobability
l→ḥ100%reflex
m→w50%reflex
m→ṣ50%reflex
n→n100%
s→l100%reflex
z→z100%