% -------------------------------------------------------- % ``The Ciphered Diary of an 19th Century Egyptologist'' % Emanuele Viterbo % for submission to CRYPTOLOGIA % Last revised: 7/1/97 % -------------------------------------------------------- % \documentstyle[12pt]{article} %\textwidth 16cm \textwidth 15cm \textheight 26cm %\hoffset -2cm \hoffset -1cm \voffset -3.5cm \renewcommand{\baselinestretch}{1.2} %\newcommand{\z}{\rule[-5mm]{0mm}{15mm}\underline{\rule[4.5mm]{10mm}{0.1mm}}} \newcommand{\hn}{\hspace{-0.5mm}} \newcommand{\hp}{\hspace{0.5mm}} \font \sime=sim %\title{The ciphered diary of Simeone Levi} %\title{The ciphered diary of an 19th century libertine} \title{The Ciphered Diary of an 19th Century Egyptologist} \author{Emanuele Viterbo} \begin{document} \maketitle \input{simdefs2} Simeone Levi was an Italian Egyptologist, who lived in Torino during the second half of the 19th century. His major work is the 8 volume Hieroglyphic dictionary \cite{vocab} for which, in 1886, he was awarded the prize of the Royal Academy of Lincei. He was the brother of the mother of my grand-grandfather, and I first heard about him when my grand mother gave her copy of the dictionary to my father. The dictionary is a lithographic copy of his handwritten manuscript and its aim was also to compare the Hieroglyphic words with the corresponding Coptic and Hebrew words, in an attempt of demonstrating their derivation from Hieroglyphic. Simeone Levi was born in 1843 in the Jewish ghetto of Carmagnola near Torino from a poor family and was the seventh of the ten children of a goldsmith. Struck by a paralysis at the age of two, he remained disabled for all his life and suffered for the limitations imposed by his handicap. Simeone got a degree in mathematics and initially earned his living by teaching mathematics. His interest for Egyptology only started at the age of 33, after attending a series of lectures by Professor Francesco Rossi, Vice-Director of the Egyptian Museum of Torino. His only other class mate was Ernesto Schiaparelli, who was to become famous for discovering Queen Nefertari's tomb. From that moment he entirely devoted himself to Papirology, having Professor Rossi as his guide and maintaining a competitive attitude towards Schiaparelli. Last year my grand-aunt Giorgina Levi (Simeone's grand-grand-niece) decided to find out more about the life of her famous ancestor \cite{vita} and initiated an historical research. Through the documents, she got in touch with the direct descendants of Simeone, Ettora and Massimo Levi. They had several papers, books and letters of their grandfather and among them a manuscript written in an unknown alphabet. They also reported that Simeone had handed down to his sons the duty of interpreting and reading the 355-page manuscript. But all attempts to decode the mysterious text failed even though it was given to examine to fairly expert, but possibly not very motivated, people. When I learned this story from my aunt, I could not resist my ancestor's challenge. At first I was only given photocopies of the front cover, the title page and the first page (Figs. \ref{pg1}--\ref{pg3}). That proved insufficient to decipher the text, but gave some hints and an idea of the difficulties to overcome: {\em i)} an apparently very large character set, possibly due to the non constant shape of the characters because of the hand writing; {\em ii)} uncertainty of the language used by such a polyglot. At a first glance the characters all look pretty much the same: some appear like calligraphic characters of the Latin alphabet, both small and capitals, some others of the Greek alphabet. The punctuation seems totally absent, but some accents appear occasionally, only on the last character of some words, supporting the hypothesis that the language was indeed Italian, as suggested by the fact that Simeone had expressed his wish that his descendants interpreted and read the text. On the assumption that the language was Italian, the word length statistics of the first page strongly indicated that he had not used a mono-alphabetic key, since the words appeared shorter than the average word length in Italian. Of course he could have also split up the words with some strange rule. Handwriting is usually hard to interpret at a first approach, unless it is one's own writing. Copying a few times the entire page was useful to became more familiar with the signs. This enabled to complete a first approximate table of the different signs. Their large number was a clear hint that the same Latin characters could be represented by different ciphered symbols depending on their context within the word. It was also possible to identify some equally ciphered words and this induced me to believe that the code memory did not extend across neighboring words. The observation of the cover page (Fig. \ref{pg1}) and title page (Fig. \ref{pg2}) was of little use, since nothing appeared to match with Simeone's name. The signs on these pages are written in a fancy variant of the standard font. In fact the same word appears both in the second line of the cover page (Fig. \ref{pg1}) and in the title of the first page (Fig. \ref{pg3}). The only line I was able to decipher was one with totally different signs. I assumed that they were numbers, and on the hypothesis that the text was his diary, they could indicate the year range 18??-1900 (the last two equal characters had to be ``00'', because they were different from the first ``1'' and Simeone died in 1913). The two unknown digits could have been ``43'', the year of Simeone's birth, but I could not go any further. I was going to give up when I received the photocopies of the entire manuscript. Some new essential elements became available. All the pages after the first one were numbered and the signs for the digits could be easily derived. The sign representing the same digit is different according to its position within the number. The upper left part of the character is constant but is followed by a sort of calligraphic {\em u}, if the digit is in odd position starting from right (units), or by a calligraphic {\em n}, when in even position within the number (see Table \ref{digits}). This enabled to establish without any doubt that the front cover contained the year range 1843-1900 and confirmed the hypothesis that the manuscript could be Simeone's autobiography. The next step was to search through the entire text for all the occurrences of numbers possibly corresponding to dates or amounts of money. The only way I could proceed was then by guessing some complete words. The historical research carried out by my aunt turned out to be valuable, since she could associate some particular facts in the life of Simeone to some of the years. Unfortunately the precise date, including day and month, was never available. Still the breakthrough came from the analysis of the dates. I wrote down and collected together all the words around the dates on the assumption that some of them represented articles, prepositions and names of months or seasons. The first words I could guess were {\em nel} \neS\lS \hp (in) usually preceding a year and {\em di} \diS\hp (of) preceding the possible name of a month. Then came all the variants {\em nell'}\spS \neS\lS\lS\apoS, {\em nella}\spS \neS\lS\laS, {\em del}\spS \deS\lS, {\em dell'}\spS \deS\lS\lS\apoS, {\em della}\spS \deS\lS\laS, for the two genders and for the words beginning with vowels in the Italian language. I was lucky to find four different months with the same last two signs, which corresponded to: {\em settembre} \seS\tS\teS\mS\bS\hn\hn\reS\spS (September), {\em ottobre} \oS\tS\toS\bS\hn\hn\reS\spS (October), {\em novembre} \noS\veS\mS\bS\hn\hn\reS\spS (November), {\em dicembre} \diS\ceS\mS\bS\hn\hn\reS\spS (December). % %\rule{0pt}{15pt} I also found \pS\riS\maS\veS\raS\spS which was preceded by {\em nella}, indicating a feminine word starting with a consonant. Since all months are masculine in Italian it could only be the season {\em primavera} (spring). Then I recognized the words {\em anno} \aS\nS\noS\hp (year), {\em il} \iS\lS\hp (the), {\em fine} \fiS\hn\hn\neS\hp (end) and {\em fino} \fiS\hn\hn\noS\hp (until), also related to the dates. This set of words was sufficient to break the code. From the four months and from {\em primavera} containing the strings {\em ra}, {\em re} and {\em ri}, it became clear that a basic character \rS\spS representing {\em r} was altered in the upper part according to the vowel which was following it (see Table \ref{conso}). Similarly from {\em novembre} and {\em nel} it was possible to derive the signs for {\em n} and {\em o}. With the above words the signs for the vowels {\em a, e, i, o} and some consonants became clear. With these few signs I started translating the first page and guessing some other words and I could gradually complete the set of signs. This representation of the vowels as diacritical signs attached to consonants was certainly inspired in Simeone by his knowledge of Sanscript. The first meaningful sentences appeared. Simeone was writing his autobiography, telling us about all his good and bad actions, his successes and failures, his friends and his enemies, his loves and his betrayals. To confuse even more the potential reader, Simeone on the third line of the first page (Fig. \ref{pg3}) quotes a sentence in a mysterious foreign language: \\ \maS\spS \peS\cuS\glieS\spS \riS\spS \siS\faS\faS\nuS\spS \ceS\cuS\muS\ceS\laS\spS \oS\spS \queS\aS\nS\ruS\spS \yS\taS\gnoS\spS \fuS\spS \aS\spS \peS\cuS\taS\gnoS\spS \peS\cuS\spS \niS\caS\geS\nuS\UpunS \\ ({\em Ma pecuglie ri sifafanu cecumucela o queanru ytagno fu a pecutagno pecu nicagenu.}) \\ I consulted various linguists in Ancient Egyptian, Sanscript, Coptic and Hebrew, but the language of this sentence still remains a mystery. I leave it as a challenge for the interested readers. \\ \vspace{5mm} \noindent By going on reading it was possible to unveil all the tricks he had used for ciphering. \begin{enumerate} \item Consonants followed by other consonants or at the end of a word take on the basic sign in the first column of Table \ref{conso}. \item The most important trick, which makes the code resist a simple statistical cripto-analysis is the way the vowels are encoded. As already mentioned, when they are following a consonant they are added on top of it. When they are following another vowel, either in the same or in the previous word (an exception to my first guess that the code memory did not extend across neighboring words), then they take on the basic sign for the lower part, while the upper part corresponds to the sign of the preceding vowel (see Table \ref{vowels}). This makes the signs for the vowels position depended. Vowels at the beginning of a word following a word ending with a consonant take on the basic sign in the first column of Table \ref{vowels}. \item The signs for the punctuation are also position dependent, since they change according to the last character in the preceding word, as for the vowels (see Table \ref{punctuation}). Punctuation does not break the memory for the vowels of Table \ref{vowels}. For example a word ending with {\em ra} followed by a comma, then followed by another word starting with an {\em o} would be encrypted as $\ldots$\raS\AvgS\spS \AoS$\,\ldots$. \item There are no capital letters. \item Apostrophes, accents and hyphens are not encrypted, but no sign is used for word hyphenation. \item Some special pairs of consonants, typical of the Italian language, {\em ch, gh, gn, gli}, were encrypted with a special sign, following the same rules of the consonants (see Table \ref{spechar}). \item The {\em qu} is a single sign, in which the {\em u} is omitted being redundant. \item Finally some words are occasionally misspelled and the spacing between words is not always well defined. \end{enumerate} The complete set of signs is reported in Tables \ref{digits}--\ref{punctuation}. Letters {\em j,k,x,y,w} do not belong to the standard Italian alphabet and appear very rarely in the text in some foreign words or names. These were among the last to be deciphered. All combinations of characters not allowed in the Italian language are only reported for completeness and are marked with an asterisk. In the last pages of the manuscript Simeone tells us that he wrote the entire manuscript between February and March 1900, while he was already immobilized at home by his paralysis. Simeone Levi died in 1913. While creating the entire ciphered font with Metafont \cite{Knuth}, in order to typeset this paper, a new interesting feature of the signs became clear: whenever a sign is ascending or descending this corresponds to an ascending or descending character in a common calligraphic alphabet. This property is a very helpful reminder when writing the ciphered signs and enables {\em real-time} ciphering once the basic rules are known. The Metafont character set for use within \LaTeX can be obtained through anonymous ftp at {\tt entropy.polito.it}. % ------------------------------------------------------------------------- \begin{table} \begin{center} \Large %\begin{tabular}{||p{6mm}||p{11mm}|p{11mm}||p{6mm}||p{11mm}|p{11mm}||} \begin{tabular}{||c||c|c||c||c|c|c||} \hline \hline & odd & even & & odd & even \\ \hline \hline 0 & \uzerS & \nzerS & 5 & \ucinS & \ncinS \\ \hline 1 & \uunoS & \nunoS & 6 & \useiS & \nseiS \\ \hline 2 & \udueS & \ndueS & 7 & \usetS & \nsetS \\ \hline 3 & \utreS & \ntreS & 8 & \uottS & \nottS \\ \hline 4 & \uquaS & \nquaS & 9 & \unovS & \nnovS \\ \hline \hline \end{tabular} \end{center} \caption{\label{digits} The digits: {\em odd} is for units and hundreds and {\em even} is for tens and thousands.} \end{table} % ------------------------------------------------- \begin{table} \begin{center} \Large %\begin{tabular}{||c||p{10mm}|p{10mm}|p{10mm}|p{10mm}|p{10mm}|p{10mm}||} \begin{tabular}{||c||c|c|c|c|c|c||} \hline \hline & & a & e & i & o & u \\ \hline \hline % & & \hspace{3mm}a & \hspace{3mm}e & \hspace{4mm}i & % \hspace{3mm}o & \hspace{3mm}u \\ \hline \hline % & & a & e & i & o & u \\ \hline \hline b & \bS & \baS & \beS & \biS & \boS & \buS \\ \hline c & \cS & \caS & \ceS & \ciS & \coS & \cuS \\ \hline d & \dS & \daS & \deS & \diS & \doS & \duS \\ \hline f & \fS & \faS & \feS & \fiS & \foS & \fuS \\ \hline g & \gS & \gaS & \geS & \giS & \goS & \guS \\ \hline h & \hS & \haS & \heS & \hiS & \hoS & \huS$^*$ \\ \hline j & \jS$^*$ & \jaS$^*$ & \jeS & \jiS$^*$ & \joS$^*$ & \juS$^*$ \\ \hline k & \kS$^*$ & \kaS$^*$ & \keS & \kiS & \koS$^*$ & \kuS$^*$ \\ \hline l & \lS & \laS & \leS & \liS & \loS & \luS \\ \hline m & \mS & \maS & \meS & \miS & \moS & \muS \\ \hline n & \nS & \naS & \neS & \niS & \noS & \nuS \\ \hline p & \pS & \paS & \peS & \piS & \poS & \puS \\ \hline qu &\qS$^*$ & \quaS & \queS & \quiS & \quoS & \quuS$^*$ \\ \hline r & \rS & \raS & \reS & \riS & \roS & \ruS \\ \hline s & \sS & \saS & \seS & \siS & \soS & \suS \\ \hline t & \tS & \taS & \teS & \tiS & \toS & \tuS \\ \hline v & \vS & \vaS & \veS & \viS & \voS & \vuS \\ \hline w & \wS$^*$ & \waS & \weS$^*$ & \wiS & \woS$^*$ & \wuS$^*$ \\ \hline x & \xS & \xaS$^*$ & \xeS$^*$ & \xiS$^*$ & \xoS$^*$ & \xuS$^*$ \\ \hline y & \yS & \yaS$^*$ & \yeS$^*$ & \yiS$^*$ & \yoS$^*$ & \yuS$^*$ \\ \hline z & \zS & \zaS & \zeS & \ziS & \zoS & \zuS \\ \hline \hline \end{tabular} \end{center} \caption{\label{conso} The consonants.} \end{table} % ---------------------------------------------------------- \begin{table} \begin{center} \Large %\begin{tabular}{||c||p{10mm}|p{10mm}|p{10mm}|p{10mm}|p{10mm}|p{10mm}||} \begin{tabular}{||c||c|c|c|c|c|c||} \hline \hline & & a & e & i & o & u \\ \hline \hline % & & \hspace{3mm}a & \hspace{3mm}e & \hspace{4mm}i & % \hspace{3mm}o & \hspace{3mm}u \\ \hline \hline a & \aS & \AaS & \EaS & \IaS & \OaS & \UaS \\ \hline e & \eS & \AeS & \EeS & \IeS & \OeS & \UeS \\ \hline i & \iS & \AiS & \EiS & \IiS & \OiS & \UiS \\ \hline o & \oS & \AoS & \EoS & \IoS & \OoS & \UoS \\ \hline u & \uS & \AuS & \EuS & \IuS & \OuS & \UuS$^*$ \\ \hline \hline \end{tabular} \end{center} \caption{\label{vowels} The vowels.} \end{table} % ------------------------------------------------------ \begin{table} \begin{center} \Large %\begin{tabular}{||c||p{10mm}|p{10mm}|p{10mm}|p{10mm}|p{10mm}|p{10mm}||} \begin{tabular}{||c||c|c|c|c|c|c||} \hline \hline & & a & e & i & o & u \\ \hline \hline % & & \hspace{3mm}a & \hspace{3mm}e & \hspace{4mm}i & % \hspace{3mm}o & \hspace{3mm}u \\ \hline \hline ch & \chS & \chaS$^*$ & \cheS & \chiS & \choS$^*$ & \chuS$^*$ \\ \hline gh & \ghS$^*$ & \ghaS$^*$ & \gheS & \ghiS & \ghoS$^*$ & \ghuS$^*$ \\ \hline gn & \gnS$^*$ & \gnaS & \gneS & \gniS & \gnoS & \gnuS \\ \hline %gl(i) & \glS$^*$ & \gliaS & \glieS & \gliS & \glioS & \gliuS$^*$ \\ \hline gli & \gliS & \gliaS & \glieS & -- & \glioS & \gliuS$^*$ \\ \hline \hline \end{tabular} \end{center} \caption{\label{spechar} Some special consonant pairs.} \end{table} % ------------------------------------------------- \begin{table} \begin{center} \Large %\begin{tabular}{||c||p{10mm}|p{10mm}|p{10mm}|p{10mm}|p{10mm}|p{10mm}||} \begin{tabular}{||c||c|c|c|c|c|c||} \hline \hline % & & a & e & i & o & u \\ \hline \hline & & \hspace{3mm}a & \hspace{3mm}e & \hspace{4mm}i & \hspace{3mm}o & \hspace{3mm}u \\ \hline \hline , & \vgS & \AvgS & \EvgS & \IvgS & \OvgS & \UvgS \\ \hline ; & \pvS & \ApvS & \EpvS & \IpvS & \OpvS & \UpvS$^*$ \\ \hline : & \dpS & \AdpS & \EdpS & \IdpS & \OdpS & \UdpS$^*$ \\ \hline . & \punS & \ApunS & \EpunS & \IpunS & \OpunS & \UpunS \\ \hline ? & \pinS & \ApinS & \EpinS & \IpinS & \OpinS & \UpinS \\ \hline ! & \pesS & & & & & \\ \hline \hline \end{tabular} \end{center} \caption{\label{punctuation} The punctuation.} \end{table} \section*{Acknowledgments} I thank Professor Ezio Biglieri for carefully reading the translated manuscript and pointing me out two words containing the letter {\em k}, which was still missing to complete the alphabet. I would also like to acknowledge Professor Michele Elia for introducing me to cryptography and for his helpful advice. \begin{thebibliography}{99} \bibitem{vocab} S.~Levi: {\em Vocabolario Geroglifico--Copto--Ebraico}, Litografia Salussolia, Torino, 1887--1894. \bibitem{vita} G.~Levi, E.~Viterbo: {\em Simeone Levi matematico ed egittologo}, in preparation. \bibitem{Knuth} D.~E.~Knuth: {\em The MetafontBook (Volume C of Computers and Typesetting),} Addison-Wesley, 1986. \end{thebibliography} \section*{Biographical Sketch} Emanuele Viterbo was born in Torino in 1966. He received his degree in 1989 in Electrical Engineering from the Politecnico of Torino, Italy. From 1990 to 1992, he was with the European Patent Office, The Hague, Holland, as a patent examiner in the field of dynamic recording and in particular in the field of error-control coding. In 1995 he received his Ph.D. in Electrical Engineering from the Politecnico of Torino. He currently holds a post-doctoral fellowship at the Politecnico di Torino, Italy. His current interests are in lattice codes for Gaussian and for fading channels, algebraic coding theory and cryptography. % ------------------------------------------------------------------------- \clearpage \begin{figure} \vspace{22cm} \caption{\label{pg1} The cover page.} \end{figure} % ------------------------------------------------------------------------- \clearpage \begin{figure} \vspace{22cm} \caption{\label{pg2} The title page.} \end{figure} % ------------------------------------------------------------------------- \clearpage \begin{figure} \vspace{22cm} \caption{\label{pg3} The first page.} \end{figure} \end{document}