The Ugaritic Cuneiform Alphabet: Structural Relationships and Incremental Design Patterns in the Sign Inventory

Second-millenium BC Eastern Mediterranean consonant–vowel (CV) syllabary inscriptions are widely attested1 . We assume that
most syllabaries were learned in order of descending syllabogram frequency. We name this principle Collapsed Syllabary Order
(CSO). This preprint investigates whether the Ugaritic Cuneiform Alphabet (UCA) was structured with regards to a CSO.
In order to compute syllable frequencies, we first gather a dataset of ancient texts in related languages. Some of the dataset has
undergone vocalic reconstruction by experts.
We also define how to compute a CSO and use our dataset to compute it under labels: Ugaritic (ULCSO), Phoenician (PCSO),
Amorite (Amorite CSO) and Arabic (ARABCSO). The CSOs are then added to our input together with attested alphabetic orders and
the descending character frequency orders of given languages (DFO).
The study can then be broken down into two parts: each candidate order is mapped onto the UCA via a character correspondence
table.
To enable quantitative analysis of the orders, we introduce a discrete structural representation of UCA signs that allow us to
measure graphical similarity between signs. This way, we compute each orders’ Average Neighbor Distance (AND). Orders in
the input are also compared using Damerau–Levenshtein edit distance as well as Spearman and Kendall–Tau rank correlation.
All measurements are evaluated against a baseline of random permutations.
Strong evidence is found for there being a link between CSOs and multiple existing alphabetical orders, opening up the potential
for better understanding of how alphabetical orders came to be and how they may relate. Finally, we discuss how results could
both shine a light on the graphical choice of UCA as well as improve our understanding of several other ancient scripts.