Skip to content


Biblia Hebraica Stuttgartensia (Amstelodamensis)

This is the text-fabric version of the Hebrew Bible Database, containing the text of the Hebrew Bible augmented with linguistic annotations compiled by the Eep Talstra Centre for Bible and Computer, VU University Amsterdam.

The text is based on the Biblia Hebraica Stuttgartensia edited by Karl Elliger and Wilhelm Rudolph, Fifth Revised Edition, edited by Adrian Schenker, © 1977 and 1997 Deutsche Bibelgesellschaft, Stuttgart.

The text-fabric version has been prepared by Dirk Roorda Data Archiving and Networked Services, with thanks to Martijn Naaijer, Cody Kingham, and Constantijn Sikkel.

There are more formats in which the data is available. In the shebanq subdirectory you find data in MQL format and in MYSQL format that directly goes into the SHEBANQ website.

In the bigTables you find ways to export the complete data as one big table, and store it in R format or in Pandas format. The notebooks bigTablesP and bigTablesR show you a few things that you can do in R and Pandas.


The source data resides on a server of the ETCBC, managed by Constantijn Sikkel. He makes that data available as an MQL database dump, together with supplementary data files. From there it is transported to this GitHub repo by means of a pipeline.

This dataset contains several versions of the BHSA, from 2011 till now. When you navigate to a version, you'll see more information about that version and its provenance.

For all versions the pipeline has been followed. For the newer versions, starting with 2016, additional data is available in other repositories. See the footer of this page.

In text-fabric it is easy to load the features of several datasets in one session.


We have compiled a list of references to give an impression by what principles and methods the ETCBC has carried out its text analyses.


The pipeline above is complicated and not free of cruft. It would be better if the ETCBC could deliver its core data directly in text-fabric format, with inclusion of the lexical features, the ketiv-qere data and the paragraph numbers. But at least all the fine distinctions that need to be made between versions have been diagnosed and dealt with in this pipeline.

Reproducible science

We intend to follow a practice that allows for data updates on the one hand, and reproduction of old results on the other.

In SHEBANQ, there are several versions of the data and they are all frozen. Data version c is peculiar, because it was intended as a moving version, alongside the frozen version. But we have abandoned the idea, and it has become a frozen version, sitting oddly between 2017 and 2021. We preserve it, because SHEBANQ has saved queries against this version.

Frozen versions in SHEBANQ will remain there forever, and publishing queries and annotations against frozen versions will remain supported.

In particular, versions 3, 4 and 4b are here to stay. Version 3 because it is relatively old, and represents an earlier stage in the feature organization of this database. Versions 4 and 4b because queries have been published that are based on them.

These versions are also firmly entrenched in the academic record, by virtue of being archived.


This work is licensed under a Attribution-NonCommercial 4.0 International (CC BY-NC 4.0). That means:

  • You may download the data and use it: process, copy, modify;
  • You may use the data to create new software applications;
  • You may use the data for research and publish any amount of results;
  • When you publish this data or results you obtained from them, you have to comply with the following:
  • give proper attribution to the data when you use it in new applications, by citing this persistent identifier: 10.17026/dans-z6y-skyh.
  • do not use the data for commercial applications without consent; for any commercial use, please contact the German Bible Society.

How to use

This data can be processed by Text-Fabric.

See also tutorial (Hebrew) and tutorial (search).

Work based on the BHSA

Martijn Naaijer and Willem van Peursen: Parsing Hebrew and Syriac morphology using Deep Learning. Blog post Netherlands eScience Center