A standard orthography simply refers to a uniform way of spelling, which in modern times is relatively easy to enforce–for example, through reforms by authoritative institutions. A notable example is the German orthography reform of 1996, initiated through an international agreement signed in Vienna by representatives of Germany, Austria, Switzerland, and Liechtenstein. Today, any deviation from such a standard is typically regarded as a typo. However, before the advent of printing, things were quite different: books were copied by hand, and spelling conventions often varied by region. Individual scribes were often inconsistent even in their own writing, resulting in diverse orthographic practices that lacked standardization.

Bible translations and language standardization

Bible translations played a pivotal role in the standardization of languages and their spelling. Luther’s Bible translation, for instance, had a profound influence on the standardization of the written German in the sixteenth century (Visser 2018: 350). Likewise, the Swedish Bible translation commissioned by Gustav Vasa, based on Luther’s version, marked a milestone in the development of a more uniform Swedish written language (Thelander 2005: 1900). Despite some irregularities, its orthography is quite consistent for its time, demonstrating a clear effort to maintain uniform spelling throughout the text (Pettersson 2005: 160).

The Sahidic Bible may not be an exception. In this brief article, I will demonstrate how Sahidic biblical manuscripts can help trace a similar process of orthographic standardization in Sahidic texts.

Method

The corpus comprises 52 texts (mostly biblical) preserved in 26 manuscripts, dating from the fourth century up to the tenth century. The goal is to computationally measure orthographic similarity–that is, the consistency of word spellings–across these texts.

The first step was to select binary variables (features) to capture orthographic behavior in the texts. These include: ϫⲉⲕⲁⲥ vs. ϫⲉⲕⲁⲁⲥ ‘so that’; assimilation of ⲛ before ⲃ, ⲗ, and ⲣ; the use of ⲏⲟⲩ vs. ⲏⲩ; and ⲡⲉⲉⲓ/ⲧⲉⲉⲓ/ⲛⲉⲉⲓ vs. ⲡⲉⲓ/ⲧⲉⲓ/ⲛⲉⲓ, among others (12 features in total).

Next, occurrences of each feature were counted in each text. Because texts vary in length, the values were normalized into ratios. For example, in Luke (sa 1), the cluster ⲏⲟⲩ appears 92 times, and ⲏⲩ 78 times. The normalized ratios are calculated as follows:

Ratio of ⲏⲟⲩ = total no. of ⲏⲟⲩ / (total no. of ⲏⲟⲩ + total no. of ⲏⲩ) = 92 / (92 + 78) = 0.541

Ratio of ⲏⲩ = total no. of ⲏⲩ / (total no. of ⲏⲟⲩ + total no. of ⲏⲩ) = 78 / (92 + 78) = 0.459

Ratios for all variables were calculated, and the resulting matrix was analyzed using Principal Components Analysis (PCA)–a statistical method that reduces dimensionality while preserving as much variability (i.e. statistical information) as possible (Redmont et al. 2023: 119). PCA is particularly helpful for identifying patterns in large datasets. The results were then visualized in a two-dimensional Euclidean space to assess the (in)consistency of orthography across texts.

Results and discussion

We generally expect orthographic consistency within a single manuscript, as texts of a manuscript were typically copied in the same location, sometimes by the same scribe. However, the texts of sa 31 (green ellipses) show considerable variation. Similarly, the sa 1 texts (red ellipse, lower right) are not closely aligned: the Luke text is clearly separated from both John and Mark (overwritten by “sa 32, Revelation”). PCA indicates that John and Mark are closer to each other than to Luke, supporting Quecke’s (1984: 18–19) paleographic observation that they were without doubt written by the same hand.

There are also discrepancies between witnesses of the same biblical book–for instance, the John texts in sa 1 and in sa 4 (grey arrows). The dispersed distribution of the texts suggests that these texts had not adopted standardized orthography. In fact, these divergent texts were copied in the fourth and fifth centuries.

In contrast, many other texts fall into two major clusters: (1) A small cluster comprising sa 4 texts, showing intra-manuscript consistency; and (2) a dense group of texts in the upper-right quadrant, nearly all of which were copied in the sixth century onwards.

Conclusion

These findings suggest that Sahidic orthography had become largely standardized no later than the sixth century. What motivated this standardization, whether institutional, ecclesiastical, or social forces, remains a question for further inquiry, requiring integration of both linguistic data and historical context. One possible explanation is provided by Funk (2013: 540; cf. also Feder 2023: 7–8) who suggests that large-scale production and distribution of Sahidic translations “was not begun until those translations had been subjected to careful revision and editing, presumably under the control of some church-based authority (whose identity and location is unknown to us).”

References

Feder, Frank. 2023. “Old Testament, Coptic Versions of the: Update,” in: Claremont Coptic Encyclopedia. https://ccdl.claremont.edu/digital/collection/cce/id/2188/

Funk, Wolf-Peter. 2013. “The Translation of the Bible into Coptic,” in: James Carleton Paget and Joachim Schaper (eds.), The New Cambridge History of the Bible, vol. 1. Cambridge: Cambridge University Press, 536–546. https://doi.org/10.1017/CBO9781139033671.029

Pettersson, Gertrud. 2005. Svenska språket under sjuhundra år: en historia om svenskan och dess utforskande. Lund: Studentlitteratur.

Quecke, Hans. 1984. Das Johannesevangelium Saïdisch: Text der Handschrift PPalau Rib. Inv.-Nr. 183 mit den Varianten der Handschriften 813 und 814 der Chester Beatty Library und der Handschrift M569. Papyrologica Castroctaviana 11. Rome and Barcelona.

Redmond, L., D. Foucambert, and L. Libersan. 2023. “Language Corpora and Principal Components Analysis,” in: D.G. Woolford, D. Kotsopoulos, B. Samuels (eds.), Applied Data Science. Studies in Big Data, vol. 125. Springer: Cham, 117–132. https://doi.org/10.1007/978-3-031-29937-7_9

Thelander, Mats. 2005. “Sociolinguistic structures chronologically II: Swedish,” in: Oskar Bandle, Kurt Braunmüller, Ernst Hakon Jahr, Allan Karker, Hans-Peter Naumann, Ulf Teleman, Lennart Elmevik and Gun Widmark (eds.), The Nordic Languages, An International Handbook of the History of the North Germanic Languages, Volume 2. Berlin, Boston: De Gruyter Mouton, 1896–1907. https://doi.org/10.1515/9783110197068-087

Visser, Arnoud. 2018. “The Luther Bible,” in: David M. Whitford (ed.), Martin Luther in Context. Cambridge: Cambridge University Press, 350–57. https://doi.org/10.1017/9781316596715.041