¹ To the best of my knowledge, the earliest list of rules suggested for the comparison of readings in OT sources is that of Walton, B., Biblia Polyglotta, Prolegomena (London: Roycroft, 1657) 1. 36–37Google Scholar (reprinted in Wrangham’s, F. edition of the Prolegomena [Cambridge, 1828] 332–36).Google Scholar Other rules for textual evaluation are included among rules for the “correction” (“emendation”) of MT or for the “detection of errors,” such as suggested by Cappellus, L., Critica Sacra (Paris, 1650; Vogel, G. I. L. and Scharfenberg, I. O., eds., Halae, 1775–86) VI.VIII.17–20Google Scholar; Clerc, J. Le (Clericus), Ars Critica (Amsterdam: Gallet, 1697) chap. XVIGoogle Scholar; Houbigant, C. F., Notae criticae in universos VT libros (Frankfurt: Varrentrap et Wenner, 1777) CXVI–CXXIV.Google Scholar With regard to the employment of textual rules, OT scholars followed the lead of other disciplines, especially classical studies and the study of the NT. For example, Cappellus, though not the father of textual criticism, but certainly the author of the first full-scale critical analysis of the text and versions of the OT, quoted extensively (VI.XII) from H. Estienne’s textual treatment of Cicero: In Marci Tulii Ciceronis quamplurimos locos castigationes (Paris: Stephanus, 1577).

² Proceedings of the Classical Association 18 (1922) 67–84Google Scholar; reprinted in Selected Prose, ed. Carter, J. (Cambridge: Cambridge University, 1961) 131–50.Google Scholar

³ For bibliographical details on the writings of Karl Lachmann (1793–1851), see Metzger, B. M., Chapters in the History of New Testament Textual Criticism (Grand Rapids: Eerdmans, 1963) 143 n. 4.Google Scholar

⁴ The comparison of details in the textual witnesses of the OT with MT rather than another source (such as a Qumran scroll) implies no value judgment; that is, “variants” are not intrinsically inferior to their counterparts in MT. MT has been chosen as the base text because that text is the textus receptus, and as such it is complete and easily accessible.

⁵ An exception should be made for the major changes in the “proto-Samaritan” texts (4QpaleoEx^m, 4QNum^b, etc.) and the Sam. Pent., whose idiosyncracies were added to the proto-Masoretic base text. The same applies to the accretions to Daniel and Esther in the parent text of the LXX.

⁶ A major difficulty is created by the situation that in some books the textual transmission actually started before the final literary form of the composition was produced. Earlier “drafts” or editions of these books preceded the final products, and although they were not circulated as much as the final product itself, they existed in a written form, and could, in principle, still be attested by an ancient textual witness. This happened in the book of Jeremiah, where an earlier edition is now attested in 4QJer_b and the LXX. See Tov, E., “Some Aspects of the Textual and Literary History of the Book of Jeremiah,” in Bogaert, P.-M., ed., Le livre de Jérémie (BETL 54; Leuven: Peeters, 1981) 145–67.Google Scholar For further examples, see Tov, E., The Text-Critical Use of the Septua-gint in Biblical Research (Jerusalem Biblical Studies 3; Jerusalem: Simor, 1981) 293–306.Google Scholar

⁷ Greenberg, Pace M. (“The Use of the Ancient Versions for Interpreting the Hebrew Text,” vTSup 29 [1978] 131–48)Google Scholar, who reckons with the possibility of different hyparchetypal readings.

⁸ For examples, see Talmon, S., “Synonymous Readings in the Textual Traditions of the Old Testament,” Scripta Hierosolymitana 8 (1961) 335–85.Google Scholar

⁹ See n. 1 above.

¹⁰ de Rossi, B., Introduzione alia Sacra Scritture (Paraa: Stamperia ducale, 1817) 99–100Google Scholar; Porter, J. Scott, Principles of Textual Criticism with their Application to Old and New Testament (London: Simms and MacIntyre, 1848)Google Scholar; Davidson, S., A Treatise on Biblical Criticism, Exhibiting a Systematic View of That Science (Boston: Gold and Lincoln, 1853) 382–87Google Scholar; de Wette, W. M. L., Lehrbuch der historisch-kritischen Einleitung in die kanonischen und apokryphischen Bücher des Alten Testaments (8th ed.; Berlin: Reimer, 1869) 233–40Google Scholar; Loisy, A., Histoire critique du texte et des versions de la Bible (Amiens: Imprimerie générale Rousseau-Leroy, 1892) 1. 239ff.Google Scholar; Kennedy, J., An Aid to the Textual Amendment of the Old Testament (Edinburgh: Clark, 1928) 189–231Google Scholar; Smith, H. P., Samuel (ICC; New York: Scribner, 1899) 395–402.Google Scholar

¹¹ Steuemagel, C., Lehrbuch der Einleitung in das Alte Testament (Tübingen: Mohr, 1912) 72–73Google Scholar; Coppens, J., “La critique du texte hébreu de l'Ancien Testament,” Bib 25 (1944) 9–49Google Scholar; Bentzen, A., Introduction to the Old Testament (Copen hagen: Gads, 1948) 1. 94–98Google Scholar; Noth, M., The Old Testament World (trans. Gruhn, V. I.; Philadelphia: Fortress, 1966) 358–63Google Scholar; Archer, G. L., A Survey of Old Testament Introduction (Chicago: Moody, 1964) 50–53Google Scholar; Payne, D. F., “Old Testa ment Textual Criticism: Its Principles and Practice,” Tyndale Bulletin 25 (1974) 99–112Google Scholar; Klein, R. W., Textual Criticism of the Old Testament (Philadelphia: Fortress, 1974) 69–75Google Scholar; J. A. Thompson, “Textual Criticism, OT,” IDBSup, 888–91; Barth, H. and Steck, O. H., Exegese des Alten Testaments, Leitfaden der Methodik (2d ed.; Neukirchen-Vluyn: Neukirchener, 1976) 20–26Google Scholar; Deist, F. E., Towards the Text of the Old Testament (trans. Winckler, W. K.; Pretoria: D. R. Church Booksellers, 1978) 243–47Google Scholar; Würthwein, E., The Text of the Old Testament (4th ed.; trans. Rhodes, E. F.; Grand Rapids: Eerdmans, 1979) 116–17Google Scholar; Barthélemy, D. et al., Preliminary and Interim Report on the Hebrew Old Testament Text Report (4 vols.; New York: United Bible Societies, 1979) V–XXXIIGoogle Scholar; Hayes, J. H., An Introduction to Old Testament Study (Nashville: Abingdon, 1979) 80–81.Google Scholar

¹² Davidson, Treatise on Biblical Criticism, 383.

¹³ Ibid., 382–87. For similar remarks, see Hayes, Introduction to Old Testament Study, 80.

¹⁴ Volz, P., “Ein Arbeitsplan für die Textkritik des Alten Testaments,” ZAW 54 (1936) 107.Google Scholar

¹⁵ An exception should be made for Payne, “Old Testament Textual Criticism.”

¹⁶ For details, see Epp, E. J., “The Eclectic Method in New Testament Textual Criticism: Solution or Symptom?” HTR 69 (1976) 211–56CrossRefGoogle Scholar; idem, “Textual Criticism, NT,” IDBSup, 891–95. These criteria have been summarized by Epp (“Eclectic Method,” 243) as follows:

1. A variant’s support by the earliest manuscripts, or by manuscripts assuredly preserving the earliest texts.

2. A variant’s support by the “best quality” manuscripts [this criterion represents, in fact, both internal and external considerations, see below p. 436 ].

3. A variant’s support by manuscripts with the widest geographical distribution.

4. A variant’s support by one or more established groups of manuscripts of recognized antiquity, character, and perhaps location, that is, of recognized “best quality.” (Italics my own.)

¹⁷ The first scholar to use external criteria extensively was probably Davidson (Treatise on Biblical Criticism). However, some of these criteria were used already before him; see, e.g., Walton, Biblia Polyglotta.

¹⁸ Klein, Textual Criticism, 74.

¹⁹ de Lagarde, P. A., Anmerkungen zur griechischen Übersetzung der Proverbien (Leipzig: n.p., 1863) 3.Google Scholar

²⁰ The main exception to this view refers to medieval Hebrew MSS, since most of the variants contained in them developed at a late stage, sometimes in the Middle Ages themselves. See Goshen-Gottstein, M. H., “Hebrew Biblical Manuscripts: Their History and Their Place in the HUBP Edition,” Bib 48 (1967) 243–90.Google Scholar As a result, one should allow himself some form of prejudice with regard to these MSS.

²¹ Würthwein, Text of the Old Testament, 114.

²² Likewise Thenlus, O., Die Bücher Samuels (ed. Löhr, M.; KeHAT; 3d ed.; Leipzig: Hirzel, 1898) XCIGoogle Scholar; Méritan, J., La version grecque des livres de Samuel, précédée d'une introduction sur la critique textuelle (Paris, 1898) 58Google Scholar; Noth, Old Testament World, 359; Thompson, “Textual Criticism,” 888; Barth and Steck, Exegese, 23.

²³ Thus also Wellhausen, J., Der Text der Bücher Samuelis (Göttingen: Vanden-hoeck, 1871)Google Scholar passim; Katz, P., “Septuagintal Studies in the Mid-Century,” in Davies, W. W. and Daube, D., eds., The Background of the New Testament and Its Eschatology (Cambridge: Cambridge University, 1956) 199Google Scholar; Smith, Samuel, 399, reacting against Löhr (n. 22): “Where G and H show variant readings, both being grammatically intelligible, they have prima facie equal claims to attention, and the decision between them must be made on the ground of internal probability.”

²⁴ The first to make this claim was probably Walton (Biblia Polyglotta, 37; p. 334 in Wrangham’s edition): “Quae lectio cum pluribus et melioris notae codicibus congruit praeferenda est ei, quae paucioribus vel non ita accurate scriptis codicibus nititur.”

²⁵ Archer, Survey, 52.

²⁶ Preliminary and Interim Report, IX. However, the Report hastens to add: “On the other hand, in treating textual evidence, one must not count text traditions, one must weigh them.”

²⁷ Ibid., “factor 2.”

²⁸ Walton (Biblia Polyglotta, 37; p. 334 in Wrangham’s edition) was probably the first to make this claim: “Quae ex codicibus antiquioribus elicitur lectio, ‘ceteris paribus’, praeferri debet ei quae recentioribus colligitur.”

²⁹ Deist, Towards the Text of the Old Testament, 232.

³⁰ Nyberg, H. S., “Das textkritische Problem des Alten Testaments am Hoseabuche demonstriert,” ZAW 52 (1934) 242.Google Scholar

³¹ Semler, J. S., Hermeneutische Vorbereitung (Halle, 1765) 3/1. 88.Google Scholar

³² Pasquali, G., Storia della tradizione e critica del testo (2d ed.; Firenze: Felice le Monnier, 1952) chap. 4.Google Scholar

³³ Especially by those scholars who adhere to a theory of local texts (recensions); for references, see Klein, Textual Criticism.

³⁴ In the textual criticism of the NT, usually a distinction is made between two types of internal criteria (“probabilities”), recently formulated as follows by Metzger, B. M.(The Text of the New Testament [2d ed.; Oxford: Clarendon, 1968] 208–11)Google Scholar: (A) Transcriptional probabilities, such as the lectio difficilior, lectio brevior; (B) Intrinsic probabilities, such as the style and vocabulary of the author throughout the book, the immediate context, and harmony with the usage of the author elsewhere. For the distinction, see already Westcott, B. F. and Hort, F. J. A., The New Testament in the Original Greek (London: Macmillan, 1881–82) 2. 19–30.Google Scholar In the textual criticism of the OT, this distinction is usually not made.

³⁵ I do not know when this rule was introduced into OT scholarship. In the research of the NT, it has been used since J. A. Bengel (“proclivi lectioni praestat ardua”) in the “Prodromus” (1725 [173489]) to his Gnomon Novi Testamenti (Tübingen: Schramm, 1742). The “Prodromus” was not accessible to me; I quote from Epp, “Eclectic Method,” 220.

³⁶ Barthélemy, Preliminary and Interim Report, IX (factor 4). For similar formulations, see Bentzen, Introduction to the Old Testament, 97; Klein, Textual Criticism, 75; and Deist, Towards the Text of the Old Testament, 244–45.

³⁷ All illustrations of this rule are questionable; yet they illustrate its background. In Gen 2:2, “on the seventh day” of MI was probably replaced in the Sam. Pent, (and also in the LXX and Peshitta or in their parent text) by “on the sixth day,” because it was not conceivable to some ancients that God finished His work on the seventh day, implying that He also worked on that day. According to this logic, the reading of MT is original. In another case, a difficult word and a hapax legomenon, ŝobel (MT), was replaced in IQIsa^a by šûlayik (Isa 47:2).

³⁸ E.g., Bentzen, Introduction to the Old Testament, 97; and Steuemagel, Lehrbuch, 97.

³⁹ Scribal errors are found in all textual witnesses, but opinions differ with regard to their recognition. in the MT of Jer 23:33 is difficult since its use of is unprecedented. It reflects a lectio difficilior as compared with the contextually appropriate (ὑμεῖς ἐστε τὸ λῆμμα) of the LXX. Most scholars agree that the reading of MT reflects a scribal error (incorrect word-division), while the LXX reflects the original reading. However, there are no accepted views on this or other variations, as is shown by an article written in defense of MT: Walker, N., “The Masoretic Pointing of Jeremiah’s Pun,” VT 7 (1957) 413.Google Scholar Likewise, in Jer 41:9 indicates, in my view, a contextually unexplainable scribal error, while the original reading () is reflected in the LXX (48:9: ϕρέαρ μέγα τοũτó έστίν). Here, too, some scholars defend MT. See, e.g., Dahood, M. J., “Hebrew Ugaritic Lexicography I,” Bib 44 (1963) 302–3.Google Scholar If, as I believe, the afore-mentioned readings in MT resulted from scribal error, the rule of the lectio difficilior does not apply to them.

⁴⁰ A case in point is 1 Sam 1:23:

In our view, there are no external or internal considerations in favor of or against MT or 4QSam^a = LXX. Both readings are possible, or, in other words, they are equally easy. Elkanah speaks of Hannah’s oath (vs 11), which in 4QSam^a = LXX is referred to explicitly, but in MT the oath is presented as God’s word. Only one of these two readings can reflect the original text (cf. p.432), and an equal claim can be made for a change of the reading of MT to that of 4QSam^a = LXX and for the reverse change. The following readings, too, are equally “easy”:

The LXX and 4QSam^a are probably derived from , which was understood as either Ίttô or ότô.

The recognition of equally “difficult” readings is admittedly very subjective. In our view, of MT in 1 Sam 20:30 (“the son of a perverse, rebellious woman”) is as difficult as the reading of 4QSam_b: (“son of deserting maidens”). The reading of MT is linguistically difficult, while that of 4QSam_b is difficult because of its contextual implications. Only one of these readings (or a third one) reflects the original text.

⁴¹ I do not know when this rule was first applied to the OT.

⁴² Klein, Textual Criticism, 75. Similarly Löhr (see n. 22 above) XLI; Archer, Survey of Old Testament Introduction, 52.

⁴³ For example, in the following instances it is more likely that an elemeni was added as an explanation than dropped as superfluous:

⁴⁴ See the recent discussion by Royse, J. R., “Scribal Habits in the Transmission of New Testament Texts,” in O'Flaherty, W. Doniger, ed., The Critical Study of Sacred Texts (Berkeley Religious Studies 2; Berkeley: Graduate Theological Union, 1979) 139–61Google Scholar (including references to the earlier studies of A. C. Clark and E. C. Colwell).

⁴⁵ E.g., l chr 11:31:

The shorter reading of the LXX could be original (cf. 1 Sam 13:15; 14:16) with dittography in MT, or secondary (for MT, cf. 2 Sam 23:29), created by haplography. In that case, the differences would have been created by a scribal phenomenon. However, it is also possible that was omitted or added because of contextual reasons. Likewise: some scholars ascribe most or all of the minuses of the LXX in Jeremiah to its shorter Hebrew Vorlage, but in some instances a case can be made for omission by a scribal phenomenon (homoioteleuton): 27(34):5 ; 27(34):20 . Even the large minus in 39(46):4–13 is ascribed by some to homoioteleuton; see, e.g., Rudolph, W., Jeremia (2d ed.; HAT; Tubingen: Mohr, 1958).Google Scholar

⁴⁶ Preliminary and Interim Report, XI (factor 5).

⁴⁷ This phenomenon occurs frequently in similar verses (both in the immediate context and in adjacent chapters) and to a lesser extent in the parallel accounts in Samuel-Kings //Chronicles and Isaiah 36–39 // 2 Kgs 18:13–20:19.

⁴⁸ Preliminary and Interim Report, XI (factor 7).

⁴⁹ In the Preliminary and Interim Report this rule is not used very often.

⁵⁰ Brockington, L. H., The Hebrew Text of the Old Testament—The Readings Adopted by the Translators of the New English Bible (Oxford: Oxford University, 1973).Google Scholar

⁵¹ Ibid., 104.

⁵² E.g, why does the addition of ὁ δϕϑείς in the LXX of Gen 31:13 reflect a lectio facilior (thus Payne) and not the translator’s exegesis? If the latter possibility were correct, the LXX would not reflect a different reading, and its Vorlage (= MT) does not need to be evaluated on a text-critical level. Likewise, are there any solid criteria for evaluating the differences between the MT and LXX in Gen 48:15? Does the plus of the LXX in Gen 48:15 indeed reflect scribal omission in MT, and could it not reflect an interpretive addition in the LXX or its Hebrew parent text?

⁵³ See p. 432 and n. 7.

⁵⁴ E.g., Davidson, Treatise on Biblical Criticism, 385; Steuemagel, Lehrbuch, 73; Bentzen, Introduction to the Old Testament, 1. 97; Greenberg, “Use of Ancient Versions,” 148.

⁵⁵ Westcott, B. F. and Hort, F. J. A., The New Testament in the Original Greek (Cambridge/London: Macmillan, 1881–82) 2. 21.Google Scholar

⁵⁶ In a modern treatise on textual criticism, L. Bieler quoted the seventeenth-century scholar Bentley as follows: “Nobis et ratio et res ipsa centum codicibus potiores sunt. I should like to add et centum regulis.” See The Grammarian’s Craft: An Introduction to Textual Criticism (Classical Folia, Studies in the Christian Perpetuation of the Classics; New York: Catholic Classical Association of Greater New York, n, d.) 45.

⁵⁷ Critica Sacra, book VI, chap. XVI (p. 719 in Scharfenberg’s edition).

⁵⁸ Lehrbuchy 233–3A. For the modern period, see Housman’s paper (n. 2 above) and Ham, E. B., “Textual Criticism and Common Sense,” Romance Philology 12 (1958/1959) 198–215Google Scholar (with bibliography).

⁵⁹ Additional note on the relation between the textual criticism of the OT and the NT: If the above analysis is correct, scholars should probably employ different approaches in the textual criticism of the Old and New Testament. In principle this situation is not surprising, since each literature should be approached according to the inner dynamics of its own textual transmission. Traditionally, textual rules are frequently used in the textual criticism of the NT, and if these rules are used prudently and critically, their employment is more appropriate in the NT than in the OT. Differences in approach to the evaluation of textual readings in the OT and NT derive from the following:

1) Textual sources of the NT are much closer in time to their supposed original text than those of the OT.

2) Probably as a result of this situation, the extent of textual variation is much smaller for the witnesses of the NT than of the OT.

3) The text criticism of the NT reckons with literally thousands of documents, for which the use of external criteria is often legitimate and certainly necessary. In the case of the OT, a much smaller number is involved?individual MSS of the ancient versions and medieval Hebrew MSS may be excluded for this evaluation. For further details on the evaluation of readings in the textual criticism of the NT, see the two articles of Epp cited in n. 16 above, the bibliography quoted by Epp, and G. Fee, “Criteria for Evaluating Textual Readings in the Textual Criticism of the New Testament,” an as yet unpublished paper read in a symposium sponsored by the Hudson-Delaware chapter of the SBL (May 1981). That symposium, at which also the present paper was read, was devoted to the evaluation of readings in the textual criticism of both the Old and the New Testament.