Pursuing Veritas

Translated by Jacob J. Prahlow

Franz Stuhlhofer, “Der Ertrag von Bibelstellenregistern für die Kanonsgeschichte,” Seirschrift für die alttestamentliche wissenschaft 100.1 (1988): 244-61.

In 1964 Sundberg wrote, “The Old Testament of the early church, as a problem of canon, has received somewhat the treatment of a stepchild in biblical studies.” Today also, two decades after this statement, people are able to repeat this statement. The attention of Old Testament scholars has dwelt only in a limited extent on this point. Reasons for this include first the disciplinary space between Old Testament Studies and Patristics. Secondly however, there lies within the subject itself a difficulty: Christian lists of canonical books of the Old Testament often first came from the fourth century, and thus impacted the history of Old Testament Studies. Canons in the first few centuries after Christ were lacking in substance. Foremost was the Old Testament discussed in the New Testament, then came a leap to Melito of Sardis, then a leap to Origen, and with this leap we find ourselves already in the 4^th century.

To fill around such a “leaping” gap is to have also tried to determine the Old Testament canon standing behind the books of Clement of Rome, Irenaeus, Tertullian, and other inferences. But with this two complicated questions arise:

1. Which Old Testament books were used? (Here too plays the difficulty of differentiating between deliberate references and random echoes.)
2. Which inferences can one pull from quotations? (In other words, to what extent is the quotation of a book synonymous with its canonical status?)

The difficulty in answering either of these questions involves the difficult task of pulling inferences from the application of the Old Testament by the Church writers, which all depends on the measurement and stand the respective observer has. Thus, it is pointed out from the one side that there is not a single quote from a deuterocanonical writing in the New Testament. Books found during the protocanonical period were used very intensively—therefore the Deuterocanon was not included in the canon. In contrast it was argued that some citations of protocanonical books were also not found. Should one therefore think that these protocanonical books do not belong to the canon? Additionally, it is argued that are also some protocanonical books with no quotes—should one therefore think that these protocanonical books do not belong in the canon? In favor of the deuterocanon is further mentioned that they are expressly quoted in the early church—so they belong to the canon. However, they are cited only by some church writers—there are others who do not, who even explicitly disapprove of these books. But even in those that specify a list of canonical books, one occasionally finds scripture quotations from books that are not in that list – so even these were not all consistent. If in an examination one includes in addition to the protocanonical and deuterocanonical books still additional materials, the issue becomes even more complex. One finds quotations from Enoch in the New Testament—should one therefore accept that Enoch belongs in the canon? Not unconditionally, because heathen authors were also sometimes expressly quoted in the New Testament—and these have not been included in the canon. From such extra-canonical quotations it is sometimes concluded that back then there was no precisely fixed canon. In a word, the issue is complex enough that everyone can maintain his position by selecting the points that support his position.

Is it possible to bring order to this chaos of opinions? Yes, but we must employ some methodological considerations.

Seven Rules of Methodology

1. A unique quotation says, speaking on its own, that the canon quotation is not yet above the cited books. If the standing of a certain book is unclear, we must verify exactly how frequently these books were quoted, from which we can know that they were canonical.

2. It is not certain that all canonical books were clearly quoted more than other books. From time to time an author gave a list of biblical books, but never quotes some of the books appearing on it and instead quotes some of the books missing from his list, sometimes even as scripture. Consider for example Athanasius: According to the indices of Muller and Thomson, he never used II Samuel, Obadiah, Nahum, Zephaniah, Ruth, Nehemiah, and I and II Chronicles. Nevertheless these books appeared in his list. On the other hand he quotes from Tobit, Sirach, and Wisdom with a formula, how he normally used the “canonical” books. Neither can the manner nor the frequency of a quotation be equated with “canon citation.” At any rate to indicate but a consideration of the frequency of use, the particular books which had practical significance were actually authoritative books (=canonical). Besides direct pronouncements (canon lists or explicit judgments of particular books)—if possible—one should have also considered indirect pronouncements to receive a greater breadth of impressions.

3. In engaging the frequency of use we must actually respect the intensity of use, and also set the frequency in relation to the volume of a particular book. For example, Clement of Alexandria used IV Ezra just as frequently as Haggai. If one considers the volume of neither of these books, one recognizes that there are plenty of quotations from Haggai, but very few from IV Ezra. The necessity of also including the particular volume in analysis was already emphasized in 1919 by Johannes Hanel: “It must also depict the absolute figure that one may always see the volume in relationship to their books.”

From the frequency of the quotes in relation to the volume of the particular book, we must determine the percentage-share of the Old Testament books compared to the volume of the complete Old Testament (Table 1). Since the early church writers almost exclusively used the Greek translation of the Old Testament, the Septuagint (LXX), we must take this as our basis. The decisive edition of the LXX is that of Alfred Rahlfs.

I have taken all the Old Testament books of the Catholic Bible as a basis, counted together with the protocanonical and deuterocanonical books. This means neither a “concession” nor an anticipatory result. In the end, it doesn’t really matter how one derives the basis. Given that IV Ezra and Enoch play a role in many church writings, I have also calculated their volume relative to the Catholic Old Testament. Both books are not preserved in Greek, which makes difficult the comparison on their volume with regard to the Greek LXX books. I have used the German and English translations and compared that volume with the German translation of Old Testament books.

If one is involved with the volume, one can conclude that (according to the Catholic usage) the Apocrypha or (according to the Protestant usage) the pseudepigrapha (also for instance Enoch or IV Ezra) would be used the right way. Though here we must also take into consideration that some registers do not consider these books at all or not at length (for instance, because of the contents of these books in the constructed registers were not as common as the contents of the canonical books). Such extracanonical books were most likely used in Alexandria, but there they were also not very intense. If we calculate the “ratio numbers” (ratio value = usage abundance in percentage divided by the amount in percentage), the rations are as follows: With Barnabas, 0.3 for Enoch and IV Ezra (here this is but a single parallel); 0.1 for Enoch in Origen; 0.1 for IV Ezra in Clement of Alexandria. No church writers of the first three hundred years reveal a corresponding use of citation volume. With this Hölscher’s assertions are questionable, “that the Christianity of the first two centuries recognized and appreciated the Apocrypha,” and “that it uses and analyzes the Apocrypha as authoritative writings.”

4. The assessment of the usage-intensity is to ensure great numbers (underlying to place/set). The total number of quotations from the Old Testament should be at least 50, but 100 would be better. But even including a total number of 100 one must still be careful: for example, someone once quoted Obadiah. This was a quote (out of a total of 100 Old Testament citations) that is 1% of citations. But because this little book makes up only 0.087% of the volume of the Old Testament, the division of citation percentage results in citation percentage ratio of 11 to 5. Hence one can be apt to conclude that the writer was concerned with his favorite book. Here we must be careful: You can see in this case how much a single occasional quote can account for, and an isolated quote could also be quite accidental. Perhaps the register-creator was wrong or the parallel given as a quotation was unintended by the writer (or that he wanted from the outset to simply indicate a parallel without claiming that is must be here a deliberate allusion), or maybe the writer had read or heard a sentence from the letter shortly beforehand. To come to safer statements, it is suggested that such books (which were probably together because they stood together on a roll) were calculatingly disseminated together. For example, one would combine the Minor Prophets; for depending on whether one, five, or ten quotes were taken from a total of 100 citations, one could conclude a slight change (0.2 percent), from the appropriate volume (1.2%), or a high ratio of use (2.3%). Such a conclusion already has more certainty.

5. As a precaution, for every church writer several registers should be evaluated. This is because each register can have its own characteristics. To illustrate this, I give several example tables for the use of the Old Testament by Irenaeus based on four registers (Table 2).

It seems here that in the Biblia Patristica (abbr. BP), that Genesis is highly overrated over against all other registries. This is a peculiarity which is noted not only in Irenaeus, but also for the other Church writers. This peculiarity then shifts the overall result, as the entire Pentateuch appears overrated and other books are somewhat undervalued. If someone does not compare other registers with the Biblica Patristica, and has to rely on only the Biblia Patristica, he may come to statements with respect to the early Church use of Pentateuch (and especially the book of Genesis) that are not true. Such characteristics were sometimes not even conscious for the creator of the register, and are therefore not always expressly stated.

Since I specified here the tables of Irenaeus in such detail, it is certainly good to also mention how Irenaeus is representative for the church of the first centuries. Over again the other Church Writers (in general), Irenaeus’ use of the Writings and the History books was a little weak, while his use of the “major prophets” was somewhat strong.

In the evaluation of registers it is also important to note that no single “quotation” terminology exists. That is, depending on how closely or how far the term “quote” (or “parallel”) is taken, it can also result in varying strengths of figures. For example, according to Henry Shires there are 239 quotations from the Old Testament expressed in the New Testament, but according to Barthelemy there are 250 formal quotations. The number of deliberate or unconscious uses will certainly be a little high, but not very high. If you are limited to such strong parallels where actual use could arise there are hardly many parallels to deuterocanonical books. Let us consider the parallels in the Nestle-Aland: If you limited your look to the quotes (“citations”), this results in a total number of 483 parallels, including 3 deuterocanonical parallels. The ratio for the deuterocanonical books then makes a total of 0.04. On the other hand, incorporating the potential quotes (“alleged”), there are 4104 parallels, including at least 339 from the deuterocanon—a ratio amounting to 0.5, and a considerable difference. It is simply not true that the parallels consistently increase as the individual Old Testament groups increase.

To stay faithful to my principle of “comparing several registers” I looked at the Nestle-Aland Greek New Testament. The first (and second) editions had a broad parallel-terminology and contained 2,753 “parallels,” of which 90 were from the deuterocanon. The third edition defined parallels in a more narrow sense and gave only 360 parallels—not one from the deuterocanon. Within the third edition the ratio for the deuterocanon naturally amounted to 0, but at any rate in both of the first editions it was 0.2. Here also we recognize the same shift. Moreover I do accept that the broader the parallel-term becomes, the more dangerous their regulations become and the more subjective judgments of the register-creator have a part to play. If Sundberg has also used the Nestle-Aland, and had divided the quotations and allusions to some of the extracanonical (and protocanonical) books, then this is a weakly stated method. He maintained this total of 145 extracanonical (and protocanonical) “quotations.” While this at first glance seems to support his thesis that the Church took over the “open” collection of books from the Jews (which was later limited), with science one has no such procedure to do more.

6. Books of mainly historical content have are a priori less likely to be quoted, when it comes to doctrinal issues (for example, finding out the biblical view on contemporary issues). It may be that a historical book belongs to the group of scriptures that was read many times, and is nevertheless cited rarely. This is true of New Testament books, especially for the Acts, and in the Old Testament where there are many more historical books. This is also observed in the use of some deuterocanonical books, namely I-II Maccabees, Tobit, Judith, and also possibly the additions to Esther. Where such books have a ratio of about 0.3 (which is otherwise rather little), that may well indicate (based on the volume) corresponding use.

7. When asked about the position of the deuterocanon, it is too undifferentiated to simply use the protocanonical writings to determine the total compared to that of the deuterocanon. Let us consider an extreme and thus particularly impressive clear example. The Loeb edition of the Works of Philo has parallel sections in 3 books from deuterocanon and 2456 parallels in the protocanon—the ratio between protocanonical and deuterocanonical books thus amounts to 819:1. This means that there is an enormous emphasis on the protocanonical books, especially when you consider the moderate volume ratio between the two groups amounts to only 5.6:1. Therefore one could argue that for Philo the protocanonical books were used much, much stronger than deuterocanonical books (once we see the fundamental problem, that we have no single clear example of Philo quoting from a deuterocanonical book). A closer examination of Philo’s use of Scripture shows that he almost exclusively used the protocanonical books of the Pentateuch (about 97% of the time). The remaining protocanonical books appear, then, not quite so overemphasized compared to the deuterocanon.