Daitch-Mokotoff Soundex Algorithm Extraction

Terminology. A phoneme is a contrastive unit of sound. The IPA symbols used below denote, among others: ʃ ("sh"), tʃ ("ch"), dʒ ("j"), ʒ (the "s" in measure), and x (the voiceless velar fricative, as in Scottish loch or Hebrew chet). Place of articulation is where the vocal tract is constricted; manner of articulation is how airflow is shaped; voicing is whether the vocal folds vibrate. See ../GLOSSARY.md for fuller definitions.

Algorithm Overview

Daitch-Mokotoff Soundex (D-M Soundex) was developed by Gary Mokotoff and Randy Daitch in 1985 specifically for Jewish and Eastern European surnames. It addresses limitations of American Soundex for these names. Each extracted rule carries a weight in [0, 1] expressing the residual edit cost of treating the two spellings as equivalent.

Key Differences from American Soundex

Feature	American Soundex	Daitch-Mokotoff
Code Length	4 characters	6 digits
First Letter	Retained literally	Encoded
Multiple Codes	No	Yes (branching)
Slavic Sounds	Poor	Excellent
Hebrew Sounds	Poor	Good

Encoding Rules

D-M Soundex uses a more extensive mapping with context-sensitivity:

Vowel Codes

Pattern	Start	Before Vowel	Other
A, E, I, O, U, Y	0	-	-
AI, AJ, AY	0	1	-
AU	0	7	-
EI, EJ, EY	0	1	-
EU	1	1	-
OI, OJ, OY	0	1	-

Consonant Codes

Pattern	Start	Before Vowel	Other
B	7	7	7
CH	5,4	5,4	5,4
CK	5,45	5,45	5,45
CZ, CS, CSZ	4	4	4
D	3	3	3
DRZ, DRS	4	4	4
DS, DSH, DSZ	4	4	4
DT	3	3	3
DZ, DZH, DZS	4	4	4
G	5	5	5
H	5	5	-
K	5	5	5
KH	5	5	5
KS	54	54	54
L	8	8	8
M	6	6	6
MN	66	66	66
N	6	6	6
NM	66	66	66
P, PH, PF	7	7	7
R	9	9	9
RZ, RS	94,4	94,4	94,4
S	4	4	4
SCH	4	4	4
SH	4	4	4
SCHTCH, SCHTSCH	2	4	4
SHTCH, SHCH, SHTSH	2	4	4
ST, STRZ, STRS, STSH	2	4	4
SZCZ, SZCS	2	4	4
SZ	4	4	4
T	3	3	3
TCH, TTCH, TTSCH	4	4	4
TH	3	3	3
TRZ, TRS	4	4	4
TSCH, TSH	4	4	4
TS, TSZ, TZ, TTZ	4	4	4
V	7	7	7
W	7	7	7
X	5,54	54	54
Z	4	4	4
ZDZ, ZDZH, ZHDZH	2	4	4
ZD	2	4	4
ZH, ZS	4	4	4

Extracted Rules

Polish Consonant Clusters

[id: 3000, name: "szcz dm", weight: 0.10, group: dm_clusters, ipa: "/ʃtʃ/"]
szcz -> ʃtʃ;

[id: 3001, name: "strz dm", weight: 0.10, group: dm_clusters, ipa: "/ʃtʃ/"]
strz -> ʃtʃ;

[id: 3002, name: "strs dm", weight: 0.10, group: dm_clusters]
strs -> ʃtʃ;

[id: 3003, name: "stsh dm", weight: 0.10, group: dm_clusters]
stsh -> ʃtʃ;

[id: 3010, name: "cz dm", weight: 0.10, group: dm_cz, ipa: "/tʃ/"]
cz -> tʃ;

[id: 3011, name: "cs dm", weight: 0.10, group: dm_cz]
cs -> tʃ;

[id: 3012, name: "csz dm", weight: 0.10, group: dm_cz]
csz -> tʃ;

[id: 3015, name: "sz dm", weight: 0.10, group: dm_sz, ipa: "/ʃ/"]
sz -> ʃ;

DRZ/DRS/DZ Patterns

[id: 3020, name: "drz dm", weight: 0.10, group: dm_drz, ipa: "/dʒ/"]
drz -> dʒ;

[id: 3021, name: "drs dm", weight: 0.10, group: dm_drz]
drs -> dʒ;

[id: 3022, name: "dz dm", weight: 0.10, group: dm_dz, ipa: "/dz/"]
dz -> dz;

[id: 3023, name: "dzh dm", weight: 0.10, group: dm_dz]
dzh -> dʒ;

[id: 3024, name: "dzs dm", weight: 0.10, group: dm_dz]
dzs -> dʒ;

ZDZ Patterns (Word-initial)

[id: 3030, name: "zdz initial dm", weight: 0.10, group: dm_zdz]
zdz -> ʒdʒ / #_;

[id: 3031, name: "zdzh initial dm", weight: 0.10, group: dm_zdz]
zdzh -> ʒdʒ / #_;

[id: 3032, name: "zhdzh initial dm", weight: 0.10, group: dm_zdz]
zhdzh -> ʒdʒ / #_;

TCH/TSH Patterns

[id: 3040, name: "tch dm", weight: 0.10, group: dm_tch, ipa: "/tʃ/"]
tch -> tʃ;

[id: 3041, name: "ttch dm", weight: 0.10, group: dm_tch]
ttch -> tʃ;

[id: 3042, name: "ttsch dm", weight: 0.10, group: dm_tch]
ttsch -> tʃ;

[id: 3043, name: "tsch dm", weight: 0.10, group: dm_tch]
tsch -> tʃ;

[id: 3044, name: "tsh dm", weight: 0.10, group: dm_tch]
tsh -> tʃ;

TRZ/TRS Patterns

[id: 3050, name: "trz dm", weight: 0.10, group: dm_trz, ipa: "/tʃ/"]
trz -> tʃ;

[id: 3051, name: "trs dm", weight: 0.10, group: dm_trz]
trs -> tʃ;

CH Variants

[id: 3055, name: "ch to kh dm", weight: 0.15, group: dm_ch, ipa: "/x/"]
ch -> x;  // Hebrew/Yiddish sound

[id: 3056, name: "ch to tsh dm", weight: 0.15, group: dm_ch, ipa: "/tʃ/"]
ch -> tʃ;  // Slavic variant

RZ/RS Patterns

[id: 3060, name: "rz dm", weight: 0.10, group: dm_rz, ipa: "/ʒ/"]
rz -> ʒ;

[id: 3061, name: "rs dm", weight: 0.15, group: dm_rz]
rs -> ʒ;  // In Polish context

Diphthongs

[id: 3070, name: "ai dm", weight: 0.10, group: dm_diphthong]
ai -> aj;

[id: 3071, name: "aj dm", weight: 0.10, group: dm_diphthong]
aj -> aj;

[id: 3072, name: "ay dm", weight: 0.10, group: dm_diphthong]
ay -> aj;

[id: 3073, name: "ei dm", weight: 0.10, group: dm_diphthong]
ei -> aj;

[id: 3074, name: "ej dm", weight: 0.10, group: dm_diphthong]
ej -> aj;

[id: 3075, name: "ey dm", weight: 0.10, group: dm_diphthong]
ey -> aj;

[id: 3076, name: "oi dm", weight: 0.10, group: dm_diphthong]
oi -> oj;

[id: 3077, name: "oj dm", weight: 0.10, group: dm_diphthong]
oj -> oj;

[id: 3078, name: "oy dm", weight: 0.10, group: dm_diphthong]
oy -> oj;

[id: 3079, name: "au dm", weight: 0.10, group: dm_diphthong]
au -> au;

[id: 3080, name: "eu dm", weight: 0.10, group: dm_diphthong]
eu -> oj;  // Yiddish pronunciation

Common Jewish Name Elements

// Gold- element
[id: 3090, name: "gold prefix dm", weight: 0.05, group: dm_element]
gold -> gold / #_;

// -berg element
[id: 3091, name: "berg suffix dm", weight: 0.05, group: dm_element]
berg -> berg / _#;

// -stein element
[id: 3092, name: "stein suffix dm", weight: 0.05, group: dm_element]
stein -> ʃtajn / _#;

// Rosen- element
[id: 3093, name: "rosen prefix dm", weight: 0.05, group: dm_element]
rosen -> rozen / #_;

// -witz/-vitz element
[id: 3094, name: "witz suffix dm", weight: 0.05, group: dm_element]
witz -> vits / _#;

// -baum element
[id: 3095, name: "baum suffix dm", weight: 0.05, group: dm_element]
baum -> boim / _#;

// -feld element
[id: 3096, name: "feld suffix dm", weight: 0.05, group: dm_element]
feld -> feld / _#;

Branching Codes

D-M Soundex produces multiple codes when a pattern can be interpreted multiple ways. In LLev, this maps to non-deterministic rules with varying weights:

// CH can be /x/ (Hebrew) or /tʃ/ (Slavic)
[id: 3055, weight: 0.15]
ch -> x;

[id: 3056, weight: 0.20]
ch -> tʃ;

Both outputs are generated; the weight difference prioritizes the more common interpretation while still matching the alternate.

Use Cases

D-M Soundex excels at matching:

• Jewish surnames (Ashkenazic, Sephardic)
• Polish surnames
• Russian surnames
• Ukrainian surnames
• Hungarian surnames
• Romanian surnames

References

1. Mokotoff, Gary & Daitch, Randy (1985). Daitch-Mokotoff Soundex System. Avotaynu, Bergenfield, NJ. (No DOI; canonical source of the D-M Soundex algorithm.)
2. Mokotoff, Gary (1997). "Soundexing and Genealogy". Avotaynu 13(3). (No DOI.)
3. Avotaynu, "Daitch-Mokotoff Soundex Coding". https://www.avotaynu.com/soundex.htm

---

← Documentation Index