ObConlang: Is anyone else besides me finding the text really a treasure
trove of wordshapes? I'm expanding the vocabulary of multiple languages as
we speak.

Eugene

2010/6/16 Gary Shannon <[log in to unmask]>

> I'm going to retract the FQG = the hypothesis.
>
> Here are letter frequencies compared to letter frequencies from my
> copy of the Brown Corpus. (English in parens) Follwing that, I've
> substituted all 26 letters bassed on frequencies. This won't, of
> course, solve the puzzle, but it does suggest a good starting point.
> My next step will be to do letter contact analysis which should expose
> which letters stand for vowles and which for consonants. (See:
> _Cryptanalysis for Microcomputers_ by Caxton Foster. pp125-132)
>
>
> 109 F 17.3% (E - 12.5%)
>  81 Q 12.9% (T -  9.2%)
>  64 A 10.2% (A -  8.0%)
>  47 J  7.5% (O -  7.6%)
>  39 V  6.2% (I -  7.3%)
>  39 G  6.2% (N -  7.1%)
>  35 E  5.6% (S -  6.5%)
>  34 I  5.4% (R -  6.1%)
>  29 O  4.6% (H -  5.4%)
>  20 T  3.2% (L -  4.1%)
>  17 W  2.7% (D -  4.0%)
>  16 N  2.5% (C -  3.1%)
>  14 S  2.2% (U -  2.7%)
>  11 P  1.7% (M -  2.6%)
>  11 D  1.7% (F -  2.3%)
>  9 Z  1.4% (P -  2.0%)
>  9 M  1.4% (G -  1.9%)
>  9 B  1.4% (W -  1.9%)
>  6 Y  1.0% (Y -  1.7%)
>  6 K  1.0% (B -  1.5%)
>  6 C  1.0% (V -  1.0%)
>  5 X  0.8% (K -  0.7%)
>  5 U  0.8% (X -  0.2%)
>  4 H  0.6% (J -  0.2%)
>  3 L  0.5% (Q -  0.1%)
>  2 R  0.3% (Z -  0.1%)
>
> Substitution strictly by freq, broken into 5-letter groups for readability.
>
> IETMA IAIET WAFAD KARID AOGLA SOIN
> ETNOH NARCA OWEET NVETE TNON
> TRZNR LERET YACSO BUIPT HETA
> TGLES EPMEL ETBSE SIETN FAITA R
> UISET VNADT HDSAC OWOET YACSO
> ERACH MOEEI ADSCN ELOEI ETAIN AEHA
> ETYAC SOETH AETVR ETEQS BTEAH VSA
> FEBIN LOETH TUARO NUSIF CAOEW O
> FAHUS LIRDT AOENR JTETU ISOER E
> ETNDS CONRU ASETN HJPPA RETWA R
> ETYCS OETHA GNETS EREPD LHETC GEI
> IERED TAAUA RLERE TMTOI OWOET TUOEI
> ETNON TAREI ETNHK CTARN RDREP SON
> ETHAU EANEN EHHFS LDRKS AB
> GLASO CAETN MECKO DTAEF SESOE
> ETENR AOETH AOALU ISOET UAOEN YEGT
> IETIQ ESEHE TNDRA OIJPI PEETU IEAS
> ETWAH DERAT ROVNA DUIEA SETUI TO
> IETHA LEWRE TTOFH MTGTE THFAL
> TBUHI PTHET AOWOE TTUOE TNGHI LO
> ETERI ESVIL ER
> ETNIZ IDOIE TYACS OETHA EREEG HLLOE
> HMAOK ALOIM HSEN
> ETFAU NMNUJ TCERE TNQSI DA
> ETSMA N
> LFNET TOETS MNR
>
> There are letter groups that repeat (OWOET, OETHA, ETHA) and words
> that almost repeat (OWOET->OWEET, SOETH->LOETH) that suggest words, or
> parts of words (since the word boundaries are unknown).
>
> Given the left-justified original Arabic seems less likely than
> English as the original source.
>
> --gary
>