Tischendorf's 8th edition Greek New Testament
with morphological tags
Based on G. Clint Yale's Tischendorf text and on Dr. Maurice A. Robinson's Public Domain Westcott-Hort text
Edited by Ulrik Sandborg-Petersen
This text and its analysis are in the Public Domain. Copy freely.
The present work is Tischendorf's 8th edition of the Greek New Testament, augmented with morphological tags, Strong's numbers, and lemmas.
Even though I am designated as the editor, the bulk of the work in the preparation of this text was done by two other men, namely G. Clint Yale and Maurice A. Robinson. Thus they deserve most of the credit for the existence of this work. Clint Yale provided the base Tischendorf text, while Dr. Robinson provided a fully parsed and lemmatized Westcott-Hort text (with some errors that were mutually corrected in the preparation of this edition). I heartily thank them both.
Preparation of the text
Clint Yale has published two Tischendorf texts. The first was published in the Public Domain on the Internet in 1997, and only contained the text — without diacritics, punctuation, or apparatus. Mr. Yale's second Tischendorf text was published later, and contained both diacritics, punctuation, and Tischendorf's apparatus.
The basis of the present work was originally Mr. Yale's Public Domain 1997 Tischendorf text, since most of the analysis was carried out using that text. However, during the last stages of preparation of the text, Mr. Yale very graciously permitted me to distribute, in the Public Domain, an accentuated version based on his later Tischendorf, for which I am very grateful. The Greek NT community owes him a debt of gratitude for this generosity.
The text has been corrected (though not thoroughly checked) against a facsimile copy of Tischendorf. The text thus mostly conforms with the printed Tischendorf. Even in cases of clear typographical errors, the text has been retained as it was printed. As Mr. Yale notes in the introduction to his later Tischendorf edition, the printed version was “typographically challenged”. Having dealt with the text in detail, I can only confirm Mr. Yale's judgment on this account.
Preparation of the analysis
Westcott-Hort and Tischendorf's 8th edition share a large percentage of common text. Therefore, the decision was made to base the morphological analysis and lemmatization on Dr. Robinson's Public Domain Westcott-Hort text.
A computer program was written to port over as much as possible of Dr. Robinson's Westcott-Hort analysis, with manual analyses being added where necessary. Only about 10740 words could not be ported over directly. Of these, only about 900 words needed manual analyses, while about another 300 word-forms were merely differences in spelling. An analytical lexicon totalling about 290 word-forms was developed for those forms which were peculiar to Tischendorf, or which needed other special attention. For the rest, an analytical lexicon was constructed automatically from the Westcott-Hort text, which was then utilized in giving parses and Strong's numbers to forms which had a unique analysis in that lexicon.
After the analysis was complete, numerous consistency-checks were made on the analysis. Grammatical relations such as agreement were checked using the linguistic search engine “Emdros”, with subsequent manual checking and correction of the cases where agreement had been broken by the process of porting the tags over from the Westcott-Hort text. After this, all instances of ambiguity in either the lemma or the parsing of a word were checked, numbering about 1030. In checking these instances, linguistic searches were again run in cases that were not attributable to genuine morphological ambiguity. If it turned out that there was a mistake, it was corrected, and the search was run again to ensure that the error was gone. After that, searches were run on all individual parts of speech (except verbs and nouns), listing all unique forms, their Strong's number, and their lemma. The surface forms were compared to the lemmas, and irregularities were weeded out.
Whenever a word-form was found which seemed not to be correct Greek, it was checked against the facsimile, and corrected if necessary.
Finally, all neuter nouns which were present either as nominative or accusative, but not both, were carefully checked to ensure that the correct case had been assigned.
Assignment of lemmas
Two lemmas are provided: One conforms to Strong's dictionary, while the other mostly conforms to Friberg, Friberg, and Miller's ANLEX. ANLEX represents more than a century's worth of additional scholarship compared to Strong's dictionary. This and other factors entail that ANLEX has, in some respects, a more fine-grained lemma-division than Strong's.
Now, the lemmas were added automatically, based solely on the assigned Strong's number. Therefore, in a few cases, a distinction which ANLEX makes is lost, since it was not made by Strong. One such example is H)=XOS which in ANLEX is two lemmas, one being masculine and the other being neuter. In Strong's dictionary, there is only one lemma, hence only one number, and hence, since the lemmas are based on the Strong's number, the distinction is lost.
The process was carried out with constant reference to a number of grammars and lexica, including BDAG, Thayer, Strong's, Abbott-Smith, Perschbacher, Liddell-Scott, and last, but not least, Friberg, Friberg and Miller's ANLEX. Blass-Debrunner-Rehkopf and Blass-Debrunner-Funk were consulted on occasion, as were a number of introductory grammars.
During the process described above, the editor had much pleasant interaction both with Professor Robinson and with Mr. Yale, resulting in mutual correction of our respective databases. All remaining errors are, of course, my own responsibility.
As of version 2.0 of the text, there is an extra field in the database, namely the “qere” for each word. In Hebrew Masoretic texts, there is a distinction between the “Kethiv” (that which is written) and the “Qere” (that which should be read). This distinction is present in versions 2.0 and above of this database. The Kethiv is that which is written in the printed Tischendorf. The Qere is what the editor thinks it should have been.
Most often, this amounts to differences in accentuation or diacritics. In a few cases, it amounts to a change in the word itself (e.g., Revelation 14:18, where the printed text reads TOI|S BO/TRUAS, where this editor thinks it should have been TOU|S BO/TRUAS, on account of the grammar).
For the vast majority of words, the Qere is identical to the Kethiv.
The parsing always follows the Qere, not the Kethiv.
The editor welcomes feedback and suggestions for improvement. He can be reached via electronic mail:
This text has a website:
Aalborg, April 2010
The tagging scheme is exactly the same as that used by Dr. Robinson in all of his texts. It is described in the file called “parsing.txt”, available at http://eBible.org/usfx/parsing.txt.