The Study of Textual Criticism






There are few aspects of exegesis that are more complicated and taxing than that of doing textual criticism.  It is the discipline that requires the exegete to know not only the languages, but all the information about the manuscripts and the versions as well as the scribal tendencies.  It also assumes that the textual critic will be familiar with the Bible, especially the literary characteristics and tendencies of each writer. Most students in seminaries and divinity schools are simply not prepared well enough for this work.  They could be, but curricula in these institutions has swung away from biblical and theological disciplines to an emphasis on training for a profession.  The additional emphasis is fine, and probably necessary--but not at the expense of the traditional disciplines. 


In his article entitled “The Textual Criticism of the Old Testament,” Harry Orlinsky wrote:


The past several decades have witnessed a flowering of Old Testament research under the influence largely of archeological discovery.  The Biblical lands, Iraq, Syria, Lebanon, Palestine and Egypt, coming as they did under the control of England and France, became fertile ground for the rediscovery by excavation of the Fertile Crescent of old.  And although the economy and social structure of the various parts of the Near East--as of the European powers--began to change in the twenties, thirties, and forties, so that England and France have been all but replaced by the authority of the United States and the Soviet Union, and such new political groupings as the United Arab Republic, the Hashemite Kingdom of Jordan, and Israel have come into being, with even the immediate end not yet in sight, enough archaeological work is still going on--in Israel more than in Transjordan and Iraq--to satisfy the desires of Biblical scholars, if not the needs of specialists in Biblical archaeology.


At about the same time, however, a new trend began to make itself felt in higher education on both sides of the Atlantic Ocean: the humanities and the social  sciences began to give way to the applied sciences.  The curricula of high schools and colleges generally became increasingly bereft of Latin and Greek and grammar--shades of days when a public school was sometimes called Latin school or grammar school!


The consequences for the textual criticism of the Old Testament were soon felt.  Here, on the one hand, the written and unwritten documents uncovered by archaeology were attracting the attention of the students of the Biblical world; and there, on the other hand, students of this same field of research found themselves more and more unable to handle the textual criticism of the Hebrew Bible, for they were entering and leaving their seminaries and Semitics departments with less direct knowledge of Hebrew, Aramaic, Greek, and Latin than the students of earlier decades.  We have gone a long way since Ezra Stiles, president of Yale University, himself taught the freshman and other classes Hebrew, and in 1781 delivered his commencement address in Hebrew. (in The Bible and the Ancient Near East, ed. By G. Ernest Wright [Doubleday, 1965]).                     


In the years since Orlinsky wrote his article, things have not gotten any better; in fact, the study of Theology and Bible in general have been diminished over the years.  As a result, ministers and other educators today more than ever find themselves having to make critical decisions about translations, biblical interpretations, and major decisions about theology and ethics, all with less formal training in the disciplines to do so.   All that can be done in seminaries and graduate programs is to offer a survey of what the discipline of doing textual criticism would look like should someone actually learn it.   But seminary students need this survey, if for no other reason than to make them more careful in doing what exegetical work they do.  Because trained in it or not, they will have to deal with it.


The student’s first introduction to Old Testament textual criticism is usually baffling because of the amount of information that is necessary to understand what is going on.  There is a vast amount of literature on the various texts and text-types of the Old Testament, the different versions of the Old Testament, and the critical theories about such.  The little book by Ernst Wurthwein, The Text of the Old Testament (Grand Rapids: Wm. B. Eerdmans, 1979), provides a good introduction to all this material.  It is, though, a general survey.  But that is sufficient in seminary, because the purpose in studying this part of the discipline is not to make the students into a textual critic, but to make them aware of the text of the Bible--how it was written, how it was preserved, how reliable it is, and how to think through textual difficulties.  


Thinking through textual difficulties takes the most training.  It requires a good knowledge of the languages and then of the manuscript evidence--what the variations are.  This involves being able to read the apparatus in the foot notes of the Hebrew Bible and then evaluating the readings according to the canons of textual criticism.  You will only do a little of this, enough to make you aware of the process of thinking through a textual problem, so that when you read different versions or different commentaries you will be able to evaluate some of what they are saying.


The modern expositor does not have the luxury of avoiding this issue--unless that “expositor” (if the term applies) is planning to ignore the Bible.  The different versions of the Bible in English will make it necessary for you to say something significant about the changes in the readings.  Moreover, modern commentaries often change the text in their discussions.  You will have to determine if they are warranted, or if like scribes of old they are simply the preference for an easier reading by the commentator.  This matter, obviously, is at the base of all other disciplines in biblical and theological studies.


The task of the true textual critic, then, is to uncover, identify, or restore the original text of the Bible.  This is called “Lower Criticism.”  “Higher Criticism” deals with determining the author, date, purpose, and integrity of the books of the Bible, which is quite different.  “Textual Criticism” deals with manuscripts and versions of the Bible, and not, as some today think, with interpretation methods.   The reading in the bibliography below will be most helpful for any who wish to pursue this further.


But it must be reiterated--our purpose will be to survey the information about manuscripts, versions, and scribal activities, as well as the method for doing textual criticism, so that you will have a better understanding of the Holy Scriptures and how they have been preserved and translated.







Ap-Thomas, D. R.  A Primer of Old Testament Textual Criticism.  Oxford: Basil Blackwell, 1965.


Cross, F. M. and Talmon, S.  Qumran and the History of the Biblical Text.  Cambridge: Harvard University Press, 1975.


Jellicoe, S.  The Septuagint and Modern Study.  New York: Oxford University Press, 1968.


Klein, R. W.  Textual Criticism of the Old Testament.  Philadelphia: Fortress Press, 1974.


Orlinsky, Harry M.  “The Textual Criticism of the Old Testament,” in The Bible and the Ancient Near East, edited by G. Ernest Wright.  Garden City: Doubleday, 1965.  Pp. 140-169.


Roberts, B. J.  The Old Testament Text and the Versions: The Hebrew Text in Transmission and the History of the Ancient Versions.  Cardiff: University of Wales press, 1951.


________.  “The Textual Transmission of the Old Testament,” in Tradition and Interpretation, edited by G. W. Anderson.  London: Oxford University Press, 1979. Pp. 1-30.


Thompson, J. A.  “Textual Criticism,” in Interpreter’s Dictionary of the Bible, Supplement.  Pp. 886-891.


Waltke,  Bruce K.  “The Textual Criticism of the Old Testament,” in Biblical Criticism: Historical, Literary and Textual, edited by R. K. Harrison, et. al.  Grand Rapids: Zondervan, 1978.  Pp. 47-78.


Weingreen, J.  Introduction to the Critical Study of the Text of the Hebrew Bible.  New York: Oxford University Press, 1982.







The Hebrew Manuscripts


The manuscript evidence for the Hebrew text is rather late, but very well preserved.  Before 400 B.C. there is no extant manuscript (MS) of the Hebrew Bible, and so we are left to infer scribal practices from the Bible itself and from other ancient Near Eastern practices.  What can be readily demonstrated is that scribes had a determination to preserve the text.  The text survived through all the disasters and devastations because the books were considered sacred and the scribes insisted on accurate transmission.  There was a “psychology of canonicity” which fostered a care and a concern for the preservation of the sacred writings.  For a study of similar scribal care in the ancient world, see W. F. Albright, From the Stone Age to Christianity, pp. 78-79; and K. A. Kitchen, Ancient Orient and Old Testament, p. 140.


But there was also the tendency in some scribal circles to revise the text.  They changed the script and the orthography according to literary conventions; they also changed linguistic features.  We know something of how the vocalization of Hebrew changed and can thereby discern such changes.  Moreover, the priests seem to have revised synoptic portions of the text in their teaching (compare Ps. 18 and 1 Sam. 22 in the Hebrew).  On top of all of this, there would have been the accidental errors such as dittography, haplography, and the like.


From 400 B.C. until the time of the standardization of the Hebrew Text in 70 A.D., the same tendencies continued.  The presence of a text-type among the Dead Sea Scrolls (DSS; ca. 200 B.C. to 100 A.D.) identical with the one preserved by the Masoretes (whose earliest extant MS dates to ca. 900 A.D.) witnesses to the faithful preservation of the text.  We learn something of this process of preservation from the Rabbinic tradition (Talmud, Nedarim 37b-38a).


But there was also a tendency among scribes to revise.  The Sopherim (“scribes”) were “authorized revisers of the text” according to C. D. Ginsberg (in his Introduction to the Masoretico-Critico Edition of the Hebrew Bible [New York: KTAV, 1966], p. 307).  After the return from the captivity the scribes altered the script from the old form to the Aramaic form of writing.  But more importantly, some of the more liberal scribes altered the text for both philological and theological reasons.  They modernized the text by replacing archaic forms and constructions, they smoothed our difficulties, they supplemented the text with additions and glosses from parallel passages, and they substituted euphemisms for vulgarities, altered names of false gods, and safe-guarded the divine name by substituting vowels from other forms.


The result of all these tendencies was the emergence of three different recensions of the Bible: the text preserved by the Masoretes (the textus receptus), the Samaritan Pentateuch (SP), and the Septuagint (LXX)--all three of which are corroborated by the DSS.


Gesenius demonstrated that the agreements between SP and LXX can be explained by assuming a common ancestor.  This has been confirmed by the work of Cross on Samuel, and of Gerleman on Chronicles (Cross, Ancient Library of Qumran, p. 142; Gerleman, “Synoptic Studies in the Old Testament,” Lunds Universitets Arsskrift, p. 9).  The common ancestor probably existed in Palestine at the time of the Chronicler (400 B.C.).  This “old Palestinian recension” was brought to Egypt during the fifth century B.C. and was further vulgarized before it became the base of the LXX (ca. 300 B.C.).  It also survived in Palestine and became the base of SP.[1]


It may be that the “old Babylonian recension” (Cross’s description of the text that the Masoretes eventually used) was reintroduced into Palestine about the time of the Maccabees (ca. 160 B.C.).  What is clear is that in the Gospel times there was a fluid state of recensions in Palestine.  This is seen in the use of the Old Testament in the New Testament, Rabbinic literature, and apocryphal books.


The Rabbinic testimony is that there was a movement away from plurality of recensions to a standardization of the Hebrew Text.  The rules of biblical hermeneutics, compiled by Hillel the Elder, demanded a sacrosanct text.  The evidence points to the existence of an official text with binding authority from a time shortly after the destruction of the temple in 70 A.D.   Exegetical comments and hermeneutical principles enunciated by Zechariah ben ha-Kazzav, Nahum of Gimzo, Rabbi Akiva, and Rabbi Ishmael, all presuppose a stabilized text, the Masoretic text type.  The Rabbis had settled on this conservative recension and adopted it for Judaism (conservative in that it made less changes and preserved more of the unusual and difficult things).      


The work of the scribes now changed from clarifying the text to standardizing and preserving it, even with the many archaic and difficult forms in it.  Because the scribes sought to conserve the text, no further developments of any significance occurred.


The work of conserving the text was the interest of the schools of families of Jewish scholars.  They represented symbolically the vowels and liturgical cantillations by diacritical marks.  They were known as “Masoretes” (Hebrew masar means “to hand down” by tradition); their tradition is called “Masorah”; and the text they preserved and vocalized is called the “Masoretic Text.”  The work of the family of Ben Asher of the school of Tiberias (a city on the western shore of Galilee) achieved prominence with the support of Maimonides in the eleventh century.


For a summary of the history of the printing of the Hebrew Bible, see N. H. Sarna, “Bible: Text,” in Encyclopedia Judaica 4 (1971):831-35.



Ancient Versions


The Septuagint.   Wurthwein provides a basic introduction to the Old Greek Old Testament.  To that outline I shall add some helpful background from other sources.  In general, we may say that the Torah or Law was translated into Greek between 295--247 B.C., the Prophets were translated before the end of the third century B.C., and the Hagiagrapha by 132 B.C.


Lagarde argued, apparently convincingly, that all extant manuscripts of the Old Greek translations go back to three recensions mentioned by Jerome, namely, the Egyptian by Hesychius, the Palestinian by Origen, and the Syrian by Lucien.  These three in turn go back to the original Greek version.


The two modern editions of the Greek are based on Lagarde’s theory and model.  The Cambridge Septuagint, containing the Pentateuch and historical books, presents Vaticanus (Codex B)  because it is the purest.  Gaps are filled in with Alexandrinus and Sinaiaticus.  It includes an immense Greek critical apparatus.  The Cambridge Septuagint has been preserved in a Greek-English Old Testament (Zondervan) that many check first as a time saving measure.  The other edition, Rahlf’s, is the Gottingen LXX; this is a critical issue, but generally comes back to B.  It does not have the Pentateuch and historical books.  Rahlf’s edition covers these parts.


Recensions of the Septuagint.   Waltke summarizes this thorny problem:


From his studies in Samuel--Kings, Cross concluded that the original LXX was revised no later than the first century B.C. toward a Hebrew text found in the Chronicler, some Qumran MSS, quotations of Josephus, the Greek minuscles boc2e2, and in the sixth column of Origen’s Hexapla, which is not Theodotionic but also Proto-Lucianic. This so-called Proto-Lucianic recension was then revised to the kai ge revision in favor of the Proto-Masoretic Text.  The third revision came in the second century A.D. by Aq. (Aquila) and Sym. (Symmachus), who revised the kai ge recension toward the Rabbinic Masoretic Text.  Barthelemy, on the other hand, contended that this Proto-Lucianic text is the original LXX, and thus envisions only two subsequent revisions.  But G. Howard contended that both these lacked definitive proof.


Waltke then demonstrates that the evidence in the Minor Prophets more conclusively shows such a revision of the Old Greek to the Proto-MT.  From it are the recensions of the second century A.D.  Aquila, a student of Aqiba, made a literal translation to suit his exegetical principles.  Symmachus tried more for the Greek idiom.  Theodotion’s version superseded the original translation in the editions of the LXX.


In the third and fourth centuries A.D., then, recensions of Hesychius, Origen, and Lucian appeared.  Origen’s fifth column of the Hexapla was influential on later copies of the LXX.  It was a text consistently corrected to the Hebrew textus receptus and therefore most corrupt.


Waltke concludes, “In the light of this history, Lagarde is perfectly correct in saying that, other things being equal, the Greek reading deviating from MT should be regarded as the original LXX.”


The Aramaic Targums.   The Aramaic translations of the Old Testament are less helpful for textual criticism.  They were standardized later in history, but more significantly they are paraphrastic in nature, containing haggadic material, modernizations of names, explanations of figurative language, etc.   Some of the Targumic materials are helpful, though, in understanding the official interpretations of the passage in the Synagogue.   This often has some bearing on the textual difficulty.  For a discussion of the Targums, see Wurthwein’s general introduction.


The Old Latin and the Vulgate The Old Latin is probably a Jewish translation based on the LXX.  The evidence for it is based not on a complete manuscript of it, but from manuscripts exhibiting a pre-Vulgate text, quotations in the Fathers, and marginal annotations in the Vulgate.


The Latin Vulgate was commissioned by Pope Damasus for Jerome (345-420).  Jerome attempted several approaches of revising the Latin texts, and eventually worked out a translation from the Hebrew Text.  His translation of the Psalms that is in the Vulgate (the so-called Gallican Psalter) was essentially drawn from the Hexapla.


Later, under the influence of the Jewish scholars at Bethlehem, Jerome produced a translation of the Psalms into Latin that is based on the Hebrew Text, called Psalterium iuxta hebraeos hieronymi, PIH.


The Syriac Peshitta.   This was the translation begun in Edessa, started in the first century A.D. for the Pentateuch, and completed by the end of the fourth century.  It appears to follow the Hebrew closely, but may have been translated from the LXX.  The Pentateuch, Isaiah, the Minor Prophets, and in part the Psalms, show the influence of the LXX.  Ezekiel and Proverbs are close to the Targums.  Job is literal.  Ruth is midrashic.  Chronicles is partially midrashic.


The division of the Syrian Christians into Nestorians and Jacobites meant separate versions of the Peshitta (“simple”) based on the earlier translation.






In addition to a knowledge of the history and development of the texts and versions, textual critics must be aware of the kinds of changes that have been made in the manuscripts if they are going to evaluate the problems.  The following outline will familiarize the readers with the scribal tendencies, but will not provide a thorough discussion of each.  For that, see the bibliography, but especially Klein, pp. 76-82, where these samples are discussed.


Unintentional Changes


1. Confusion of similar lettersAt times a textual difficulty arose because a scribe confused a letter in the reading.[2]  Note the difference in I Samuel 14:47:


MT:      “He pronounced (them) wicked” ( y r sh y ‘ )

LXX:    “He was victorious” (y w sh ‘ -- reading a w for an r )


2. Confusion of words that sound alike The scribe may have not heard the pronunciation of the word distinctly and mistook it for another word.  I Samuel 28:2 shows such a change:


MT:      “you”  (’attah)

LXX: “now” (apparently reading ‘attah)


3. Omission because of similar endings (homoeoteleutonThe eye of the scribe may have skipped from one ending or a word or a sentence to a similar one later on, leaving out the intervening material.  Observe how this happened in the MT of I Samuel 13:15:


MT:      “And Samuel arose and set out from Gilgal to Gibeah of Benjamin”

LXX: “And Samuel arose and set out from Gilgal-- and went on his way; but  the rest of the people went up after Saul to meet the soldiers.  Then they came from Gilgal--to Gibeah of Benjamin.”


4. Omission because of a similar beginning (homoeoarchton).  This is the same kind of error as the last, although less frequent.  The scribe’s eye may skipped from one beginning to the next, leaving out the intervening material.


5. Haplography or single writingThis refers to the single writing of two letters or words which appear together, but also to the accidental omission of letters or words.  I Samuel 17:46 has such a case:


LXX: “I will leave your corpses and the corpses of the Philistine army” (the words apparently coming from consonants p g r k)

MT      For the words in italics the MT only has one p g r.


6. Dittography or double-writing.    At times the scribe would copy over again some of the words that he had just finished.  A good example comes from the text of II Samuel 6:3-4:   


“And they made the ark of God ride on a new cart, and they took it away from the house of Abinadab which is on the hill.  Uzzah and Ahio, sons of Abinadab, guided the--new cart, and they took it away from the house of Abinadab which is on the hill.”


That this is a dittography is supported by 4QSama and the LXX.


7. Incorrect word division.  The difficulty is more common in Greek manuscripts than in Hebrew because of the spacing.  The following example, however, is one which depends on reading the letter  h  as a suffix or as an article.


MT:      “And he built the city” (I Chron. 11:8  [w y b n   h ‘ y r])

LXX:    “And he built it a city” (II Sam. 5:9 [ w y b n h   ‘ y r])


8. Incorrect vocalization.  The vowel points record the traditional pronunciation, but at times this was missed, either by the Greek translators who were working from a manuscript without vowel points at all, or by the Masoretes themselves who may have mispointed it.  Psalm 130:4 has this problem:


MT:      “there is forgiveness that you might be feared” (tiwware’)

LXX:    “law” (the translator saw the consonants and assumed it was the common noun t w r ’  [tora], rather than a very rare, irregular verb he probably did not know).


9. Transposition of words or letters (metathesis).   The scribes at times got the letters reversed, changing the sense, as in I Samuel 17:39):


MT:      “and he endeavored unsuccessfully” ( w y ’ l  )--an awkward reading!

LXX:    “and he exerted himself” (apparently reading w y l ’ )


10. Substitution of synonyms.  The scribe’s memory may have accidently slipped as he put in a similar, perhaps more familiar word for the precise one.  I Samuel 10:25 has:


MT:      “each man to his home”

LXX:    “each man to his place” (also in 4QSama)


11. Assimilation of the wording in one passage to the slightly different wording in the context or in a parallel passage.  I Samuel 12:15 has:


MT:      “the hand of Yahweh will be against you and your fathers.”


The reading “fathers” is difficult.  LXXL has “your king.”  S. R. Driver suggests that the frequent use of “fathers” in verses 6-8 may have led to the change accidently.  In this kind of problem the exegete should be alert to frequent patterns and stylistic devices in the context.


12. Mistaken inclusion of marginal comments into the text.  S. Talmon (in Textus 4 [1964]:118) has illustrated this with Isaiah 24:4:


MT:      “the heights with the land (mourn)”

1QIsa    “the heights of the land (mourn)”


Talmon shows that above the line in the Qumran scroll a scribe wrote the word ‘m (if pointed ‘am, then “people”).  He thinks this was part of an alternate form of the line: “the people of the land (mourn).”  At a subsequent copying of this manuscript, the interlinear word was inserted into the text where it was thought to be the preposition ‘m (pointed ‘im), “with,” giving rise to the strange reading in the MT.


Intentional Changes


            There were scribes who occasionally felt compelled to correct what appeared to

them to be corruptions in the text.  The most reliable scribes tried to preserve the text even if they thought there were archaic or incorrect forms--but some were not so reserved in their work.


1. Changes in spelling or grammar.   Scribes who felt free to change the text tended to smooth out readings, making verbs agree grammatically with their subjects, for example.  Several minor additions might also be added to make a better, clearer reading.  Modern translations often do this as well, occasionally putting the additions in italics, but not always.


2. Harmonizations.   Scribes may add things to the text to harmonize the line with other indications in the context.  I Samuel 20:5 may be a good example; the context of verses 34-35 tells that David hid for three days.


MT:      “Let me hide in the open country until the third evening.”

LXX:    “Let me hide in the country until evening.”


3. Conflation of variant readings.   A scribe may include both variants without realizing only one was the original.  In the following verse, Ezekiel 1:20, the italicized words are missing in some Hebrew manuscripts, LXX, and Syriac.


MT:      “Wherever the spirit wanted to go, they went, wherever the spirit wanted to go, and the wheels rose along with them.”


4. Filling out names and epithets.   There are many rather involved textual problems with names in the Old Testament.  Scribes tended to give the fuller spellings of names, which in turn often led to conflated readings as well.  The following sample from II Samuel 3:3a is a thorny one:


MT:      “Chileab of Abigail ( k l ’ b    l ’ b y g l )  the widow of Nabal the Carmelite”

LXX:    “Dalouia the son of Abigaia the Carmelitess”


The other bits of evidence for this problem are as follows: I Chronicles 3:1 has the name as D n y ’ l  (“Daniel”);  the Latin Vulgate has “Cheloab”; the Syriac Peshitta has “Chelab”; Josephus has “Danielos”; and the DSS 4Q has  d l w y h  as does the LXX.


Either the boy had two names, or there has been a confusion by dittography.  One may posit an original name “Daniel” (laynd) as represented in I Chronicles 3:1.  Then, by dittography   d / l ’ b  crept into the text and  n y ’ l  dropped out or was replaced by the repeated writing.  This would explain the name “Dalouia” in the LXX and Qumran.  Then, the   d   in   d l ’ b  changed to k perhaps when the letters were written similarly in the Hasmonean period.  This gave rise to the MT reading of Chileab.


5. Supplying subjects and objects.   When the original text failed to mention explicitly the subject or the object, scribes tended to clarify them for the reader.  Wellhausen formed the rule that “if LXX and MT differ in respect of a subject, it is probable that the original text had neither” (see Driver, Notes on the Hebrew Text, p. lxii).


6. Expansion from parallel passages.   There were times when the scribe was familiar with passages elsewhere and that caused him to add to his text from the familiar parallel section.


7. Removal of difficult expressions.   This section pertains to matters of history, geography, or theology that seemed to the scribe incorrect or offensive.  One example is Job 1:5, 11 and 2:5, 9, where the expression “curse God” was the original.  The expression was offensive, and so was changed for the euphemism “bless God”--although the scribes knew it was “curse God.”


8. Replacement of rare words with more common ones.   Scribes might have a tendency to use a more familiar term in the copy.  Isaiah 39:1 seems to show such a case:


MT:      “[Hezekiah] became well” ( w y kh  z  q )

1QIsa   “[Hezekiah] became well” ( w y kh y h )


These samples provide a brief survey of the tendencies of the scribes in copying and translating manuscripts.  The selection of samples is in no way designed to prejudice the reader’s bias towards MT or LXX in any problem.  The scribal tendencies listed here are true of all copyists, ancient and modern.  Knowing what scribes do helps the exegete work through the problems in the Hebrew Bible with a good deal of confidence; consequently, in most cases the resolution of the problem will be fairly clear.









When working through a passage in the Bible the expositor is bound to come across several textual difficulties.  These will be obvious from the different ways that English Bibles translate the verse, or from the discussions in the commentaries.  Those that are major textual problems will have to be studied, in so far as the expositor is equipped to do that.  This may mean relying on the better commentaries for the discussion.


In the Hebrew Bible, however, the data is laid out pretty clearly in the footnotes (called the apparatus) and in the marginal notes (known as the Masorah, the traditional ideas of the Masoretes).  Knowing how to use this material is the easiest, quickest avenue to resolving any textual problem. 


Footnotes in the text of the Scripture direct the reader to the apparatus for the information.  There the material will be arranged verse by verse for easy reference.  A series of abbreviations and signs will indicate what versions and manuscripts read some variant.  If there is no footnote on a given verse, then the expositor may safely assume that in all the scores and scores of manuscripts in the various languages there is no variation.  It is amazing how few textual problems there are in the Old Testament passages, given the number of manuscripts in all the languages!  People sometimes think the verses are just loaded with problems.  But, for example, in the Pilgrim Psalms, Psalms 120-134, there may be one, two, or three major textual problems in each of the fifteen psalms, and these can be settled with confidence by following sound principles.[3]  Occasionally there are verses that have problems on several of the words..  But the text for the most part is in excellent shape.  And where there is a textual problem, it is not a question of whether or not we have the original, it is a question of which it is.


Sometimes the editors of the Hebrew Bible put suggestions in the footnotes; these are called conjectural emendations--how they think the text should be changed.  There is no manuscript evidence for these; they simply reflect what modern scholars think the text should have.  Conservative textual critics will read these, but hardly ever accept them since there is no evidence for them whatsoever.  Some of the more critical commentaries (like the old ICC series) adopted most of these and set about re-writing a good deal of text.


Not all textual difficulties are worth studying.  One has to develop a sense or an instinct for this (which is why we say exegesis is both a skill and an art).  If the variants are minimal, or make no difference in meaning or translation, they need not be studied.  For example, if the noun is singular with a collective meaning, and a version makes it plural, this would be of no serious consequence.  Or, if the variant has hardly any evidence--say only one late version, then this too may be passed over most of the time.  As a general rule, if there is manuscript evidence, and the variant makes a difference, then it must be looked at closely.


When you survey English Bibles or the Prayer Book you will be able to tell right away where the major ones are.  And whether you are good at textual criticism or not, you will have to be able to say something about the difference.


The method for doing textual criticism involves external evidence, internal evidence, and intrinsic evidence.  For those who can work with the languages the process will be complete and much easier.  For those who must rely on secondary sources, being aware of the process should help in discerning whether a commentary is following sound procedure or not.  Too often commentators, like more liberal scribes, want to smooth out the text to where it makes better sense to them.  That may not be the right thing to do.


External Evidence


External Evidence refers to the assessment of the manuscripts and versions that have the various readings--which are older, which are better.  Rather than simply having a chart of the major manuscripts and mechanically adding up the witnesses on each side, Old Testament textual work requires a little more individual evaluation.  It is true that the Masoretic Text (MT), our Hebrew text, is rugged and reliable, and usually the better reading.  But at times its readings may be too rugged and may not preserve the best reading (meaning, a word, a phrase, part of a word).  It is true that the Greek (generally referred to popularly as the LXX or Septuagint) is an inferior text type on the whole; but it may preserve the superior reading here or there and so cannot be swept aside in a cavalier fashion.  Other versions will have their bearing on the consideration as well, and each alignment of witnesses (manuscripts and versions) must be given careful evaluation. 


Even then, however, a decision is not normally made on the basis of external evidence alone.  Do not assume that what the MT has is correct and that it would require a mountain of evidence to dislodge it.  The Babylonian Text Type (that is, the proto-Masoretic text) was selected as the authoritative text because it was superior in the Torah.  It is clearly not superior in some of the other books.


The most important step in dealing with a textual problem is understanding the problem.  One cannot evaluate the evidence until all the variants have been translated and understood--how they interpret the verse.  Each has to be translated literally and analyzed carefully for grammar and syntax.  This should be sufficient to isolate the differences and identify the difficulty.  It will be seen that problems usually grew up where the Hebrew text was difficult, archaic, or rare. 


The apparatus of the recent Hebrew Bible will put the actual Greek forms from the Greek translation in the footnote, as well as the Syriac and the Aramaic.  The process involves understanding what these say (by translating and analyzing them), and then by attempting to discover what Hebrew form they might have been looking at to get what they have.   Older Hebrew Bibles (Kittel’s edition) and better commentaries, will try to reconstruct the Hebrew Vorlage for you, but not always.  Once you have put the form in Hebrew, then you can tell how it differs from the standard Hebrew text, if it does.


There are some shortcuts and time savers that expositors learn about in the process.  For example, there is a copy of the LXX available with an English translation (as you will see in some of the assignments).  This lets you see immediately the difference in the Greek.   But two cautions are in order here.  First, make sure you compare that English of the Greek to a fairly literal modern English translation of the Hebrew (like the New American Standard Bible) to see most clearly the difference.  And second, note that that column Greek-English Bible is simply one Greek manuscript--Vaticanus (Codex B).  It is undoubtedly the best, but may not always be correct.  So before relying on it completely, make sure that the Greek it is translating into English for you is the original Greek reading.  This can be done by checking the discussion in the commentaries.  


So once you lay out the two or three possible readings you will be able to evaluate them.  For example, in Psalm 127 the MT has “whose quiver is full” [of arrows that represent children].  The Greek has “whose desire is full.”  You cannot solve this without getting the Hebrew word for “quiver,” and then trying to see what Hebrew words the Greek translator might have been looking at to get “desire.”  You would find in the dictionary that there is a word for “desire” that has the same letters as the word for “quiver,” but not in the exact same order.  Now it will begin to look like there as a confusion of letters--but by whom?  That is where the next step will come in.  But this has to be laid out clearly first.


Evaluating the external evidence enables the exegete to make a preliminary judgment on the direction of the decision.  With the understanding of the problem in mind, an experienced textual critic will find it difficult not to be doing the internal evidence at the same time.  But the following suggestions should help the beginner to think through external evidence:


1)                  Remember that there are basically three text types--the Babylonian which is reflected in the Masoretic Text (our Hebrew Bible), the Palestinian which is reflected in the Samaritan Pentateuch, Dead Sea Scrolls, and other various minor witnesses, and the Egyptian which is reflected in the Old Greek, what we call the Septuagint.  Great caution must be exercised in sorting these families out, because sometimes there is later development in each toward the other.  For example, one has to be sure that the exact Old Greek is uncovered (and this involves comparing the Cambridge Septuagint with Rahlf’s), and then the later recensions have to be checked (in Fields) to see, for example, how Aquila changed the Greek to bring it into conformity with the authorized Hebrew text that had been established.  Actually, the whole process takes the shape of detective work, tracing clues to see where they lead.


2)                  The evaluation must consider the relative value of all these sources and the quality of them from book to book--so there is no simple answer.  It will take some time to learn this, but the Hebrew has been preserved well in certain books and not so well in others, and the Greek has been done better in some than in others (not all the same translators).  Generally, if the Hebrew and the Greek agree, that is fairly strong evidence for the original reading.  But if they have different readings, then one must consider all the evidence.


3)                  The other versions will have an important bearing on the information.  The Vulgate usually supports the MT, the Old Latin usually supports the Greek, and the Syriac may reflect a Palestinian tradition or the MT or the Greek--it should not be passed off too lightly in the material, although often it shows signs of smoothing the text.  The Dead Sea Scrolls provide a great amount of detail, and on the whole attests to the antiquity of the proto-Masoretic tradition--something the critical scholars denied for the longest time.  The Targum is of little help for textual criticism, except that it gives the understanding of the verse that was the accepted Jewish interpretation for the synagogue.  Most often it follows a paraphrastic translation rather than literal.




Internal Evidence


Once a preliminary evaluation has been made on the problem, the exegete must evaluate the matter internally.  This is essentially where the scribal tendencies are considered in the process of reasoning through what might have happened to come to a conclusion.  To be sure, this process might seem to be more subjective; but if it follows the accepted canons (rules) of textual criticism, the whole procedure can be  consistently maintained.  Unfortunately, translators and commentators often go against the canons of criticism and accept what they think is a better reading (i.e., it makes more sense to them).


So this step requires a knowledge of the kinds of errors that were made, either accidental errors or intentional changes.  It is not always easy to tell if the change was intentional or accidental.  A scribal change could be considered unintentional if we mean that the translator of an unpointed (no vowels) Hebrew manuscript made a choice for the translation on the basis of his knowledge of Hebrew and Greek, thinking that the word in question was a common word (that he knew well) but which in fact was a rare word (which he could never have known without some oral tradition).


It is hard to imagine the difficulty we would be in if it were not for Jewish oral tradition.  They preserved the text so accurately by memory that all the rare and difficult forms were preserved equally as well as the common and well-known.  Thus, the Masoretes were not inventing vowels or making up words (as the Greek did on occasion), but were inventing signs and marks to preserve what had been given to them.  For the reliability of this tradition, see James Barr, Comparative Philology and the Text of the Old Testament.


When it comes to deciding on the correct reading, the textual critic is both scientist and artist--the decision will depend on both the knowledge of the data and how to use it as well as instinct and skill in understanding the work of the author.  The skill can only be developed with practice, and only on the basis of the knowledge gained by study.  Most students of the Bible will not develop the skill sufficiently to do this kind of work properly. 


But they should at least be familiar with the canons of textual criticism so that they can think clearly about the history and preservation of the text.  


1.         Where the Hebrew manuscripts and the ancient versions agree, it may be concluded with confidence that the original reading has been preserved.


It is always possible that a new manuscript will be found that might have a different reading, and it might be the preferred reading.  But after all this time with the thousands and thousands of manuscripts in the different languages it seems more likely that we have seen the variations.  Most of these variants can be evaluated with confidence.   And given the number of words, phrases, and verses that have no textual variants, there is no reason to question the preservation of the text.


2.                  Where the Hebrew manuscripts and the ancient versions differ, one should choose the reading that most readily makes the development of the other reading(s) intelligible.


This means that the exegete must work through the change from both perspectives.  For example, if the Greek and Hebrew differ, one must work through the problem starting from both sides.  So, if the Hebrew were the original reading, what would have prompted the translator of the Greek to come up with the form he did (was there an accidental change, an intentional change, and if so which).  Then, if the Greek were the original, what would have prompted the scribe in the Hebrew tradition to make the change that is there.


Here is where one finds that the reading that is the most difficult is to be preferred.  This is not automatic, for sometimes there is a corruption that is too difficult.  But on the whole, a difficult form, an unusual spelling or grammatical use, or a difficult idea, prompted some scribe or translator to make a change.  Scribes do not tend to introduce difficulties into the text!  They do not change clear forms into archaic forms or easy grammar into difficult or common idioms into unusual expressions.  So we know that the easier reading is likely to be secondary.


Of course, this requires that the one doing the textual criticism know what an easy and a difficult form would be.  That comes with study of the Bible and especially the Hebrew text.  Usually a working knowledge of beginning Hebrew grammar is enough to know what are the difficult forms and what are common.


Often the preferred reading will be the shorter reading.  This does not apply to places where the scribe deleted lines by jumping to a similar ending, and so is not automatic.  And it is of no help when the only difference is the vowels in a word.  But scribes did tend to add and clarify the text.  That has to be kept in mind.



Intrinsic Evidence


All of this is getting to the third aspect, the intrinsic evidence.  This is where the knowledge of the author of the book or passage has to be considered--and this comes in time with careful exegesis of the larger sections.  But knowing what David, or Isaiah, or Malachi would normally say, or perhaps what they nowhere else say, might influence a decision on a particular textual variation.






Ruth 1:21


Text and Apparatus.  Ruth 1:21 in the MT says, “the LORD has testified against me, and the Almighty has dealt bitterly with me.”   


The reading in question in the Masoretic Text is a ‘ n h   b y  a (‘anah  bi)  which is translated “has testified against me” (the qal perfect of this verb followed by this preposition means “answer against” in a legal sense). 


BHS under 1:21a-a in the apparatus at the bottom of the page tells us that the G(reek),  probably the  S(yriac) and the V(ulgate) attest to a variant reading.  The evidence from the Greek translation is in the rendering etapeinosen me, translated something like “he has oppressed me” or “he has afflicted me” or “he has humbled me.”


External Evidence.   The evidence, then, shows two different verbs entirely.  On the one side we have the Hebrew MT tradition, and on the other side the LXX tradition.  Normally, before we could weigh this evidence, we would have to know how well the text of Ruth was preserved in the MT and how good the translation of Ruth is in the LXX.  On the basis of that information, we could make a preliminary judgment in favor of the MT, for the Greek has several difficult places in its translation of Ruth.  But since we would not dismiss the variant reading on external evidence alone, we must think through the internal argument.


Internal Evidence.   Before reasoning through the canons of criticism we have to look for explanations of the Greek variant.  Here the exegete must go to the Hebrew dictionaries and word books and look for a word that is similar to ‘anah but has the meaning of “oppress” or “afflict.”  The dictionary by BDB is most helpful as it lists all related forms together.  Doing this one would find that there is a verb , ‘innah, a piel form, that means “afflict.” 


The translators of the Greek Old Testament were working with an unpointed Hebrew text, so they would not see any vowels that would indicate which verbal stem this was.  But they had the parallel “had dealt bitterly with me” and assumed this was to be synonymous parallelism to that.   So then, we can explain rather easily how the Greek version came about.


But if the original had the piel form, meaning “afflict,” and made such a nice parallelism between the clauses, there would be no reason why a Hebrew scribe would change it to the qal and introduce the legal terminology.   The qal in the MT  had to be the original.


So an unpointed text would explain the origin of the Greek translation; moreover, the MT is a slightly harder reading.  This is why we can conclude that the MT preserves the preferred reading.        



Ruth 2:7


Text and Apparatus.  Ruth 2:7 says, “and she came and remained from the morning until now--except she sat in the house . . . . ”


This last clause in the MT   bzeh shibtsh habbayit,b is “except her sitting in the house.”   The verb form is the qal infinitive construct of yashab with the 3fsg suffix.  


BH under  2:7b-b says: G(reek) is:   ou katepausen en to agro.


BH3 under 2:72-2 says : l (egendum) c(um) G(reek)--“read with the Greek.”  Then the editors reconstruct what Hebrew the Greek was looking at:  lo’  shabetah bassaddeh;   it also adds what the V(ulgate) must have read: welo’  shabah habbayit , and finally it tells you that the Syriac deleted the whole clause ( > ).


External Evidence.   In this example I have pieced together the information from both editions of the Hebrew Bible that are available.  The BHS text is what most people buy today; but BH3 has more material and arranges it slightly differently.


The first thing to do is to translate the signs and abbreviations in the apparatus.  Once this has been done the variants must be translated and the Hebrew reconstructed.  This is where BH3 comes in handy, for it tries to reconstruct what the Greek translator was translating--BHS simply gives the Greek form.


The evidence would look like this:


1)                  Masoretic text has “except she sat in the house,” reading with the qal infinitive construct of yashab, with a pronominal suffix.  The reading raises the question of what house she might have sat in for a while.


2)                  LXX has “she has [not] stopped in the [field] a little,” reading apparently a qal  perfect of shabat.   Note that BH3 says to read with this reading as the better reading (they do this a lot).  Apparently the Greek translator saw the Hebrew letters   sh b t  h and immediately thought of “she rested” rather than “her sitting.”


3)                  Vulgate has “not for a moment has she returned to the house,” apparently taking the verb form as shabah (< shub), “return.”


4)                  The Syriac text deleted the entire clause.


We should note in passing that the versions all had different resolutions for the translation of what in the Hebrew text was a difficult infinitive construct form.  We should also note that the Vulgate and the MT mention “the house,” but the LXX has “field.”  Because the versions are not strongly united against the Hebrew in these two readings, one would hesitate before making a change in the MT here on the basis of external evidence alone.


Internal Evidence.  Regarding the reading of the verb, the MT is to be preferred because the form  shibtah  best explains the origin of the other reading(s).  If the MT had had the common word shabetah, “she rested”--so well-known in the Old Testament--there would be no reason why a scribe would change it to the rare and difficult infinitive.  And if the Greek translator saw the unpointed sh b t h in the text, he would most likely have thought of shabat,  “to rest,” especially in this story.  The Vulgate was also trying to make sense out of it to reflect the best idea.


Regarding the reading “house,” the MT is to be preferred again.  The reading is difficult because it raises the question of what house she could have stayed in for a little while.  The LXX tried to harmonize this with the idea of working in the field--but that is not supported by the Vulgate which also reads “house.”  Rudolph suggested that there might have been a dittography:   shabeteh/habbayit to account for “house.”  However, it could also be that the word was omitted in the LXX for a similar reason, and agro introduced.


People who worked in the fields had little shelters that they could use for shade and rest.  The use of the term here need not imply that there was a normal “house.”






[1] Be aware that the term “Palestine” here is more convenient than accurate.  The land was not called Palestine until 135 A.D. when Hadrian named it that and banned the Jews from Jerusalem.

[2] The term “reading” refers to any variation under consideration, a word, a part of a word, a phrase, a whole line.   If the apparatus in the Hebrew Bible offers no variant reading, one assumes all the versions and manuscripts agree (they may not, but this is a safe assumption).

[3] For example, in Psalm 125 there are three variants (and I list them so you get a better idea of the scope of what we call a variant or a reading): In verse 1 the MT has “Those who trust in the LORD are like Mount Zion which can never be moved but remains forever; Jerusalem--as the mountains are around it . . .” and the Greek has “Those who trust in the LORD are like Mount Zion; the one who remains in Jerusalem can never be moved.”  The second problem concerns the MT reading of verse 3, “the rod of wickedness” which the Greek has “the rod of the wicked.”  And the third problem is that the MT has that as the subject, “The rod of wickedness will not rest on the righteous” but the Greek has “The LORD will not permit the rod of the wicked to rest on the righteous.”  In all three cases the Greek was smoothing out a more difficult (and therefore correct) Hebrew text.