Diversity and Unity in the Literary Structure of Genesis

David W. Baker


A.R. Millard & D.J. Wiseman, eds., Essays on the Patriarchal Narratives. Leicester: IVP, 1980. Hbk. ISBN: 0851117430. pp.189-205.
[Reproduced by permission]
To read the other essays, click here.


Within the body of a text there are indications of internal divisions, especially on the higher discourse levels such as paragraph, story or episode. While these indications fall into a limited number of categories, individual examples are extremely numerous. These indications will be better understood in their context than presented in a list. This study will deal with the text as it appears at one period in time, namely the Hebrew/Masoretic text of the book of Genesis. This approach is a necessary preliminary to a study of the text, which some scholars believe is itself derived from sources. We must analyse that which is objectively determinable (the present Masoretic text) before we can study that which is subjectively proposed (the source documents). The existence of these documents is at times proposed on the grounds of the division markers in the text. To avoid a circular argument, we shall study the overt form and function of these markers before making a suggestion concerning their implications.

Scholars have already noted the importance of the study of the divisions of the text in Old Testament studies. Muilenburg, in his 1968 Society of Biblical Literature presidential address, took up Eissfeldt's call to study the inter-relationships between text sections rather than simply to multiply these sections by repeatedly dividing the text.[1] We must study the text as a literary unit to see where it is divided into smaller sections (which Kessler calls the 'macro-structure') as well as the devices used to mark the divisions and indicate the unity (which he calls the 'micro-structure').[2] Each passage must be seen in its objective Sitz im Text before it can be studied in its often more vague and subjective Sitz im Leben.[3]

Muilenburg and others who espouse this approach do


not deny that we should value the study of smaller literary sections in addition to a study of the larger units of which they are a part. One of the problems in this analysis of smaller sections of the text, however, and one which form critics also recognize, is how to determine where a literary unit begins and ends. To determine this, Muilenburg says, is the first concern in rhetorical criticism.[4]

Tucker noted that the study of the structure of literary Units is valid not only in reference to '"original" units of oral expression'.[5] A 'unit' can mean anything from the entire text down to a single word. These text units do not necessarily correspond to source-critical divisions and attributions, although the latter can enter into the study at a later stage, when they can be compared with the textual sections determined by form criticism.[6]

We can place indications of textual divisions in three general categories: (1) syntactical indications of a discontinuity in such areas as time, subject or venue, (2) structural indications of the framework of the text, including headings, subscripts and summaries as well as repeated literary patterns or formulae, (3) rhetorical devices which point to a self-contained unit distinct from its context.


That the book of Genesis is a distinct unit is shown by the indications of discontinuity which occur at each end of the text. In the very first verse a temporal discontinuity is indicated by the adverb 'in the beginning'. At the end of the book there is an implicit change of subject, since Joseph, the protagonist of the preceding chapters (Gn. 37-50), is dead.[7] These represent two types of discontinuity which indicate divisions in the text: change of time and change of subject.

1.1. Change of time

Sometimes a narrative unit is specifically stated to begin or end. When Pharoah was dreaming in Genesis 41, it is recorded after each of his dreams that 'he woke up' (verses 4, 7). Also, when the events foretold by his dreams were happening, the end of one textual unit is marked 'and the seven years of plenty were completed' (verse 53) while the next unit begins 'and the seven years of famine started' (verse 54).[8]

There are also a number of other textual divisions


which are marked by an explicit time change. This is especially important when stress is laid on the progress of time; for example, in the flood narrative. The start of the flood is marked by a marginal time reference (that is, one on the periphery of the story unit) which gives Noah's age: 'and Noah was six hundred years old' (Gn. 7:6). Following the account of the entrance of the animals into the ark, the section is concluded with a wait of seven days (verse 10). Then, the exact date of the flood is given (verse 11), followed by other indications of time which serve to mark divisions in the narrative by marking gaps, or discontinuities, in time (verses 12, 17, 24; 8:4, 5, 6,10, 13-14)

This same sort of division is seen in other literature of the annal type; for example, in the Babylonian Chronicles in which each section is headed by mu X ('the X year of'),[9] as well as in Kings and Chronicles.[10]

Andersen noted that ages are often indicated as time references marginal to a story either at the beginning or the end, or else between episodes within a story.[11] An example of this is the division between Genesis 16 and 17. In Genesis 16:16 Abram is said to be eighty-six when Hagar bore Ishmael. The next section, in which Abram is renamed and Yahweh institutes the covenant with him, starts out with a statement that Abram was ninety-nine years old (Gn. 17:1). The first of these marginal time references is of interest because it is a circumstantial clause with the inversion of the subject and the predicate 'Now Abram was eighty-six years old when Hagar bore Ishmael to Abram' (Gn. 16:16).[12] The second, 'Abram was ninety-nine years old', starts a new section concerning Yahweh's covenant with Abram. This is shown by the gap of thirteen years between it and the previous section which concerns the birth of Ishmael. Other such time references which start a new episode are Genesis 23:1; 26:34; 24:1[13] and 27:1 (referring to agedness, rather than a specific age);[14] 34:25; 37:2).[15]

This second marginal time reference is of interest because it is introduced by the verbal form way[e]hî, which Andersen and Gesenius-Kautsch noted as often marking the transition to a new textual unit.[16] This verb frequently signals a discontinuity in circumstances from the preceding section. Most of these reflect a lapse in time, as did the example above in which there was a time gap between Ishmael's birth and the following events.[17] Thus, while 'the imperfect with -consecutive serves to express actions, events, or states, which are to be regarded as temporal or logical sequence of actions, events, or states mentioned immediately before',[18] way[e]hî in this


form appears to stress the discontinuity rather than the continuity between separated passages.[19] Again it is important to consider each case in its context, since the verb can have other uses as well as that of indicating a temporal discontinuity.[20]

In addition to the absolute time indicators which separate text sections, there are also relative time markers which bind sections, as well as marking their boundary.[21] These include such phrases as 'after (this)' (),[22] 'again' ('ôd),[23] and 'also' (gam).[24]

Time changes also mark textual divisions in Akkadian texts, especially annals. For example, time is used as a division marker in Grayson, Chronicles 1, where each year's activities are separated from those of another year by a line.[25] The activities are arranged in the order of the months in which they occur.[26]

1.2. Change of subject

A change in subject matter can be an obvious indicator of a discontinuity in the text.[27] A passage giving the family tree of Noah's sons (Gn. 10) is clearly distinct from one concerned with the building of a town (Gn. 11:1-9). The same is true of passages in which different personnel are involved. On its most mundane level, this change of personnel marks sub-paragraphs such as those within a dialogue.[28] For example, in the discussion between Yahweh, the man and the woman concerning the couple's sin in Genesis 3:9-19, different text sections are shown by different grammatical subjects (verses 9, 10, 11, 12, 13 [two], 14) or indirect objects (verses 14, 16, 17), which comprise a speech formula such as 'X spoke/commanded/said to Y'.

Other larger sections can be marked by the introduction of a new character. Frequently this involves the use of a circumstantial clause, i.e. one which breaks the ordinary Hebrew narrative prose chain of -consecutive plus prefixed verb (or, more rarely, plus suffixed verb). Commonly this is done by inserting the subject, which generally follows the predicate in Hebrew prose, between the -consecutive and the verb.[29] Following the account of Yahweh's covenant with Abram in Genesis 15, a new person is introduced and a new section started by the circumstantial clause 'Now Sarai, Abram's wife, had not given birth for him' (16:1).[30] A circumstantial clause is not the only indication of subject change, however, since a new subject, or at least the resumption of a subject which has been already introduced, can follow a -consecutive verb, according to the ordinary Hebrew narrative sentence structure of -consecutive verb plus subject (eg. Gn. 4:


25; 11:1; 12:1, 4, etc.).

Andersen noted that circumstantial clauses can occur at the end of an episode, although it is at times difficult to determine whether the clause is at the beginning or the end.[31] Termination can be made clear by the introduction of a new subject in the following verse, thus starting a new section. For example, following a dialogue between Yahweh and Abraham in Genesis 18 it is stated that 'Yahweh went away when he had finished speaking to Abraham, and Abraham returned to his own place' (verse 33). The last clause is circumstantial in form and closes a section, since the following verse introduces new subjects, i.e. 'now two angels came to Sodom' (19:1).

Change of subject marks divisions elsewhere in the Old Testament as well as in other Semitic literatures. This is especially common in historiographic texts such as annals. Examples are abundant in the histories of the Israelite kings (e.g. 1 Kgs.1:1; 5:26; 10:1; 11:26 and passim) as they are in those of Mesopotamian kings; cf. E. Grayson, Chronicles, 14 (cf. p.30), 18, 20-22 (cf. p.56); L. W. King, Annals of the Kings of Assyria (1902; cited as Annals), p.41:36; 47:89; 56:73; 57:88; 70:22 (Tiglat-pileser I: 1114-1076 BC) and other royal inscriptions. Some of these texts also include indications of a time change in conjunction with that of personnel.[32]

1.3. Change of time and subject: genealogical lists

Genealogical lists combine aspects of discontinuities in time and in subject. While they occur in Genesis (e.g. Gn. 5, 10 and passim), they are by no means restricted to that book. Recently Wilson published a detailed study of the form and function of genealogies in the Hebrew Bible.[33] He has compared these with contemporary oral genealogical lists among African tribes, and with other ancient near-eastern genealogical material. His rather general conclusion is that 'genealogies seem to have been created and preserved for domestic, politico-jural, and religious purposes'.[34] However, the function of genealogies in the literary structure of the text in which they occur is mentioned only rarely, and requires further study.

An abbreviated form of genealogical data, 'A son of B (son of Y)', is commonly used in the Old Testament as a type of epithet. These were used to identify an individual more precisely. The 'epithet' form does not usually serve as a structural division marker within the Hebrew text, except incidentally when the person identified by it is encountered for the first time. A genealogy marking the end of a text section is sometimes used to relate the main figure of the preceding narrative to some other known


personnage. For example the genealogy in Ruth 4:18-22 links Boaz to the proverbial Perez in the past (cf. 4:12; Gn. 38:29) and to David in the future. While representing a change of subject, a continuity with the past is also indicated. Thus the genealogy serves to emphasize both discontinuity and continuity.

Genealogies also link narratives concerning two different time periods. For example, Noah is introduced by a genealogical link to Adam through Seth in Genesis 5:3-32, while in Genesis 11:10-28 Noah and his sons are linked through Shem to Abram and Lot. This use of a genealogy unites two groups of traditions in which a relationship is not internally clear. The basis of these genealogies is not natural descent alone, since other criteria must be used to determine the line which was to be pursued. As Williamson noted concerning the early genealogies in Chronicles, which he acknowledged as being taken from Genesis, the line followed in these genealogies is that of the elect, that is, it leads to the founding of the people of Israel.[35]

Genealogies are not necessarily restricted to the main line of biblical history, and may concern secondary lines which are not pursued beyond the end of the genealogy. Adam's line through Cain is given in detail in Genesis 4:1-24 and subsequently abandoned, while the line of Seth is simply introduced by a two generation genealogy in this chapter (verses 25-26). It is then taken up and continued for ten generations in 5:3-23, as noted above. Also, Genesis 10 takes up the lines of Noah's sons, Japheth, Ham and Shem. The order appears to be a result of listing the most important line last, with the least important, and youngest, first.[36] The lineage of Shem is then resumed, again for ten generations, in Genesis 11:10ff. Other secondary genealogies include those of the lines of Ishmael (Gn. 25:12-18), Esau (Gn. 36:2-5, 9-14) and Abraham's descendants through Keturah (Gn. 25:1-4).[37] All of these are between major text sections, except the last, which separates two episodes. In each case the genealogies mark divisions.

1.4. Change of venue

Another discontinuity which indicates a division between sections in a narrative text is a change in the location or venue of the events recorded.

Coats noted a formula which serves as a transition marker between two units.[38] From the examples which he gives, this could be called a 'settling formula', and it involves the subject taking up residence (yshab) at a given location, In sane cases this transitional formula


has no clear relationship with either the preceding or the following section. Following the genealogy of Esau in Genesis 36, 'Jacob settled () in the land of the residence of his fathers, that is, in the land of Canaan' (37:1), This serves as a bridge between the preceding genealogy and the following tô1[e]dôt formula and the Joseph story. Other occurrences of this 'settling-formula' indicate either the beginning or the end of a narrative unit.[39] There are also cases where synonyms of settlement are used with the same function.[40]

Some passages are bounded by a formula which contains a 'settling-formula' but is more complex. These have been named variously 'departure' or 'itinerary' formulae,[41] and involve a specific discontinuity in location as well as a settling down. These passages mention the departure from one place and the arrival at another. One of the more complex examples of this extended formula is Genesis 11:31 which makes a transition between the pre-history of Genesis 1-11 and the following patriarchal history when its records that 'Terah took Abram... and Lot... and they set out () from the Chaldaen Ur to go () to the land of Canaan. They came () as far as Haran and they settled (wayyesh[e]bu) there.' This is a complete itinerary from departure to settlement.[42] Other such formulae in Genesis are shorter,[43] and some of them use verbs other than to express the settlement aspect.[44] Some of these formulae do not contain the 'settlement' clause, containing only an indication of departure and arrival[45] or only one of these elements.[46] At times there is only a verb which indicates that travel is taking place.[47]

Venue change as a mark of textual division in narrative texts is by no means limited to Genesis but is common in other Old Testament narrative texts as well.[48] A form of venue shift also indicates divisions in other literatures. For example, in the Sumerian King List the material is divided according to the different locations in which kingship resided.[49] The same form of change of dynasty is also used in the neo-Assyrian Dynastic Chronicle, at least in some parts of the text.[50] In other cases, such as royal inscriptions concerning military campaigns, divisions (some of which are marked by extra-textual indicators) also have a change of venue to mark a discontinuity.[51]


At times the biblical text is divided into distinct sec-


tions by its formal literary structure. One method of this which is used in Genesis is panel-writing in which a structured set of component statements is repeated in the same form a number of times.[52] An example is found in Genesis 1 in which the pattern (1) speech clause ('and God said'), (2) command, (3) execution of command, (4) formula of divine approval, (5) ordering formula[53] is repeated eight times.[54] This panel writing is used to indicate sections in various lists such as genealogies (eq. Gn. 5; 9:28-29; 11:10b-26). It can also be seen to unify narrative passages (eg. Gn. 9:12-17;[55] 15,[56] 17[57]). Due to the fixed progression of elements within the panel, it is possible to determine the start and finish of each.

In addition to formulaic structure indicating the boundaries of a textual unit, there are also verbal formulae which can indicate the start or end of a section. One of these is the clause 'these are the tôl[e]dôt of X' which occurs eleven times in Genesis.[58] I have argued elsewhere that the formula is neither strictly a heading, starting a new section, nor only a colophon, closing a section. It appears, however, to be ambiguous, at times opening and at times closing a portion of the text.[59] Whichever way the clause refers, it divides the text at a juncture where the concern of the narrative is focused down to a smaller group of people until finally it is focused, in the Pentateuch, on Aaron and Moses,[60] or in Ruth, to David. As well as this theological interpretation of the formula, it should be noted that it always occurs in conjunction with other division markers.[61]

There are also other formulae with the same demonstrative pronouns ( 'these') which are commonly used as summary statements in Genesis, and most commonly refer to the preceding text section,[62] but at times refer to the following, as headings.[63] Therefore, while these demonstrative formulae do mark a division between text sections, it is the context which determines whether they are anaphoric (summary subscripts) or cataphoric (summary headings). It is also the context which indicates whether they mark paragraphs or larger units such as episodes or complete stories.

In addition to formulaic summaries which indicate textual boundaries, there are other different clauses which serve this function. In Genesis 50, Pharaoh gave Joseph permission to go and bring his father to Egypt '(7) so Joseph went up to bring his father. There went up with him all of Pharaoh's servants, the elders of his house, and all of the elders of Egypt; (8) also all of Jacob's house and his brothers and his father's house... (9) there also went up with him chariotry and cavalry, it was a


large group.' Here the final clause serves to summarize the list. This ends the text unit, because in the next verse a change of venue is marked, starting a new unit. Summaries are not restricted to lists, however, since the action of a whole passage can be summarized at the end. Thus, after an account of Abraham circumcising his whole household (17:23-25) we are told '(26) On that day Abraham and Ishmael his son were circumcised, (27) and all of his retainers, houseborn or purchased from others, were circumcised with him.'[64]

These division markers fall into two general categories. One set of markers is syntactical, indicating a discontinuity between two literary units. This can be a discontinuity of time (indicated by references to ages and dates, by the verb or by more specific statements of temporal discontinuity - 'after that', etc. - as well as implicit times gaps), or a discontinuity of location, indicating a movement from one place to another (along with the related 'settling-formula' and its variants), or a discontinuity of subject (in which new personnel are introduced or stressed). All three of these discontinuities can be marked by syntactical discontinuities, that is a circumstantial clause which breaks the ordinary narrative chain of consecutive verbs.

The second general category of division markers is formal, involving those summaries which can serve either as introductions to or as final sections of a literary unit. These can be either formulaic or non-formulaic in character. Into this category can be grouped explicit statements of the commencement or termination of a section. Panel-writing also belongs to this category.


In addition to these division indicators there are also marks of literary unity and discontinuity, grouped under the title of rhetorical devices, which can also delineate units of greater or lesser extent. One of these is the repetition of vocabulary or phraseology at the beginning and at the end of a textual unit. This is called an inclusio, and sets the unit apart from its context.[65] An example of an inclusio is Genesis 1:l-2:4a. In 2:4a, all of the non-formulaic vocabulary, ie., all except for the tôl[e]dôt formula itself, is a repetition of that found in the first verse, 'created, heaven and earth'. The pair 'heaven and earth' also occur in the same order in both verses. While this order is that most frequently used in the Old Testament, it is deliberately used here (as can be


demonstrated by the second half of 2:4, in which the order is reversed) to indicate a dichotomy between the two halves of the verse. This inclusio thus marks the seven day creation account as a unit separate from the following section.[66]

This rhetorical device delineates literary units a few verses (eg. 5:1b-2) to several chapters long. Coats suggested that the 'settling'-formula in 37:1, opens the Joseph story ('And Jacob settled in the land where his fathers had resided, in the land of Canaan'), and that the same formula in 47:27a closes it ('And Israel settled in the land of Egypt, in the land of Goshen'). This parallel, or inclusio, he suggests, not only defines the boundaries of the story, but also 'suggests ... a structural dialectic in the Joseph story itself'.[67]

Andersen noted that a similar rhetorical device the recapitualtion or 'echo' of some important point, sometimes through the use of a circumstantial clause - can be used to close an episode.[68] The example of this device which he gives is the story of Noah's nakedness in Genesis 9. After being told, in a series of -consecutive verbs, that Shem and Japheth went in backwards to cover their father, the action is 'echoed' in verse 23b, 'their faces were towards the back so they did not see their father's nakedness', a circumstantial clause. The new section opening with a time discontinuity in the next verse ('when Noah awoke from his wine'), shows that this recapitulation does in fact close the previous section.

The criteria derived from discourse analysis indicating a textual division is harder to control than some previously mentioned since, as McEvenue has shown, recap- itulatory 'echo' is also used to link two or more units which are separated by a division marker. An example which he cites is the time reference 'after the flood' which forms part of the inclusio in Genesis 10:1, 32, and is an echo of 9:28a and is again mentioned in 11:10. Other components of this inclusio are also echoes.[69]

Therefore, as well as showing divisions between two sections of a text, the rhetorical devices mentioned above also indicate a textual unity. The inclusio was by definition a repetition of vocabulary at the beginning and end of a text section, thus dividing the text by showing a textual unity, i.e. that portion enclosed by the inclusio. The echo can also serve as a uniting feature, since in it one passage resumes a theme or vocabulary introduced in a previous passage. The latter must either predate or be contemporaneous with the former.

Another rhetorical linking device is the chiasm or 'palistrophe'.[70] Not only can a chiasm indicate the unity


of one section in contrast to others,[71] it can also serve to unite two passages that have been shown to be distinct units into a larger whole. For example, the chiasm 'heaven and earth - earth and heaven' in Genesis 2:4 indicates a conscious unification of the two passages Genesis 1:1 - 2:4a and 2:4bff.

A rhetorical device in Genesis related to the echo that also repeats vocabulary of one section in another, is prolepsis, or the anticipation of an event or action before it is actually recorded. This occurs in the form of a summary note. For example, Genesis 6:5-7 records Yahweh's dissatisfaction with man's wickedness and his vow to destroy man. Verse 8, which ends a section by a circumstantial clause (cf. section 1.2.) tells of Noah's uprightness. After a tôl[e]dôt-formula division marker (Gn. 6:9), another circumstantial clause reaffirms Noah's good character (verse 9b) and verses 10-7:24 detail the destruction of man. This not only exemplifies the use of summary anticipation followed by detailed fulfilment, but also utilizes the chiasm A-vow to destroy B-Noah's righteousness, B-'Noah's righteousness, A'-destruction, to unite two separated sections.[72] Another interesting example is found in Genesis 37:36 and 39:1, which record Joseph's descent to Egypt and his purchase by Potiphar. The two notes of this sequence of events in these two verses are separated by the Judah-Tamar incident in Genesis 38. Genesis 39:1 is then a recapitulation and resumption (Wiederaufnahme) of the narrative which had been temporarily interrupted by the intervening chapter. The resumption indicates the unity of the passage as it now stands, including chapter 38.[73]


In addition to these structural or rhetorical indicators of textual unity there are also explicit references in the text which show the same thing. Sometimes one section is specifically referred to in another, as when reference is made to a previous occurrence of an event. This has been already noted above in the discussion of relative time references (section 1.1.). The adverb 'there' (sham) is also used to indicate a place designated earlier in the text, thus uniting the two references.[74] Another unifying feature involves questions, commands and promises. They look forward to the response in which there is an answer, obedience, or fulfilment;[75] or, at the fulfilment or answer, reference can be made back to the original question, etc.[76]


Richter also noted the unifying character of verbal and nominal pronouns. Not only do anaphoric pronouns tie a clause to the previous explicit designation of the referrent; so too do the personal elements in Hebrew verbs. These, in addition to the verbal component, also contain person, number and gender within themselves, so that a subject could be explicitly given at the start of an episode and resumed only in the verb during the course of the narrative. For example, in Genesis 21:14 'Abraham awoke in the morning and took bread and a sack of water and gave it to Hagar ... she went and she wandered in the desert (15) and she threw the child under a bush ... (16) and went and sat ... and wept ... (19). She saw a well of water, went and filled the water sack and gave the youth to drink.' The resumed third person feminine subject here unites the whole passage.[77] Related to this are those passages in which a noun is introduced in one grammatical function and then is resumed as a pronoun with another function,[78] e.g. Genesis 30:25: Jacob (subject) speaks to Laban (indirect object), verse 27 Laban (subject) speaks to Jacob (indirect object), verses 28, 29 Jacob - (subject) speaks, verse 31 he (Laban - subject) speaks.

Similar to these two indicators of textual unity are the enclitic personal pronouns which also resume previously specified nouns. There are numerous examples of these pronouns referring to something mentioned immediately before but there are also occasions in which the pronoun refers to something several verses previously. For example in Genesis 14:9 Chedarlaomer and his colleagues are mentioned by name and then resumed in an enclitic pronoun six verses later when 'he (Abram) split up against him by night, he and his servants, and attacked them and chased them as far as Hormah' (Gn. 14:15). This example is noteworthy because it binds together two sections which were divided by a circumstantial clause in verse 10.[79] Another similar example is the resumption of Abraham who is named in Genesis 17:26 and resumed in pronouns in verse 27 and 18:1-2. The division between the chapters was noted by Speiser and von Rad.[80] The anaphoric pronouns uniting these two passages are especially intriguing since critical opinion attributes Genesis 17 to P and Genesis 18 to J, the latest and earliest sources of the Pentateuch respectively. According to the structure of these chapters as they now stand, a source critic would have to conclude that someone could make reference to a piece of literature which appears five or six centuries later, a hypothesis which does not seem likely.

One further syntactical feature which can serve to unify a text is the quasi-anaphoric use of the definite


article.[81] We have already noted the resumption of definite nouns by the use of anaphoric pronouns. The same also occurs when a noun is detailed at one point (e.g. 'behold three men were standing before him', Gn.18:2) and resumed with only the definite article later on ('the men turned from there and went to Sodom', Gn. 18:22; 'the two angels', Gn. 19:1; 'the angels', verse 15; 'the men', verses 12, 16). These latter forms presuppose the explanation in 18:2. This future is common within passages attributed to one source[82] and can also indicate unity between different sections, either as determined by the criteria mentioned above, or those proposed by adherents to the documentary hypothesis.[83]


The Hebrew text of Genesis fits well into its contemporary literary milieu as far as its structure is concerned. Divisions between text sections in other documents, while at times employing overt, extra-textual indicators of separate units (e.g. rulings or spaces) which are not used in the Hebrew Old Testament, share many similarities with those divisions which are indicated in Genesis within the text itself. There is nothing out of the ordinary in the structure of the book which might indicate that it is a heterogeneous amalgam of originally separate sources which have been melded, at times leaving evidence of crude joins, as some have proposed. As far as the matters discussed in this paper are concerned, Genesis appears to be a well-structured literary document.

Concerning the absolute or even the relative dating of the final composition of this book we can determine nothing significant from this study of textual division indicators, since the criteria used to determine section boundaries are not peculiar to any specific period. We can, however, deduce something from an observation of the unifying features studied here. This paper has shown that in some cases texts attributed to E or P refer to material attributed to J. While this is not remarkable, similar references in the 'early' J to the 'late' P must be explained. This calls for a thorough re-examination of the theory of source documents, or at least of the relative dates of those proposed by classical source analysis.


[1] J. Muilenburg, 'Form Criticism and Beyond', JBL 88, 1969, pp.1-18; cf. O. Eissfeldt, 'Die kleinste literarische Einheit in den Erzählungsbüchern des Alten Testaments', Kleine Schriften I (Tübingen: J. C. B. Mohr, 1962), p.49.

[2] M. Kessler, 'A Methodological Setting for Rhetorical Criticism', Semitics 4, 1974, pp.22-36.

[3] Ibid.

[4] JBL 88, pp.8-9; cf. also G. M. Tucker, 'Prophetic Speech', Int 32 (1978), pp.32-33, and F. Schicklberger, 'Biblische Literarkritik und linguistische Texttheorie', TZ 34 (1978), pp.65-81. The designation 'rhetorical criticism' is to be preferred to 'structural analysis', since the latter can lead to confusion with the entirely different study of 'structuralism'; see Kessler, Semitics 4, p.32.

[5] Tucker, Int 32, pp.32-33.

[6] Ibid.

[7] Gn. 50:26 - 'And Joseph died at one hundred and twenty years of age. They embalmed him and put him in a casket in Egypt.'

[8] These two verses, while indicating temporal discontinuity showing a boundary between sections, are united by their literary form (verb + seven years + subject clause) as well as by a word play on the verbs (t[e]hillênâ/tik[e]lênâ) which are differentiated only by the distinction /h/-/k/ which is more graphic than phonetic. For other examples of an action being said to begin or end see Gn. 17:22 (P); 6:1; 9:20, 24; 20:8; 28:16, 18 (J); 22:3 (5); cf. Gn. 21:15 (E); 24:15, 22; 27:30 (J) in which a new section is headed by a note regarding the termination of something related to the previous section. For a discussion of the formula , see D. Irvin, Mytharion (Neukirchen-Vtuyn: Neukirchener Verlag, 1978), p.25.

[9] A. K. Grayson, Assyrian and Babylonian Chronicles (Locust Valley: J. J. Augustin, 1975; cited as Chronicles), passim.

[10] 1 Kings 15:1, 9, 33; 16:8 and passim. This is not a statement concerning the relative dates of this type of division marker, but is saying only that it is common and does serve, when it occurs, to separate two text sections.

[11] F. I. Andersen, The Sentence in Biblical Hebrew (The Hague/Paris: Mouton, 1974, cited as sentence), p.81; see also Irvin, Mytharion, p.28.

[12] See section 1.2. For other marginal time references in the form of a circumstantial clause closing an episode, see Andersen, Sentence, p.81.

[ 13] A circumstantial clause, see ibid., p.80.

[14] A relative rather than an absolute reference to a discontinuity in time.

[15] Ibid., p.87.

[16] Ibid, p.63, GK, para.111f; see also Schicklberger, TZ 34, p.70.

[17] Andersen, Sentence, p.63. See also G. S. Ogden, 'Time, and the Verb in OT Prose', VT 21 (1971), pp.451-469, especially p.462 where he notes that the ww plus prefix conjugation of the verb indicate a 'break in thought between the two clauses in question' of a longer or shorter interval.

[18] GK, para.111a.

[19] G. Coats, From Canaan to Egypt (Washington: Catholic Biblical Association of America, 1976; cited as Canaan), p.21 noted that way[e]hî sometimes does not separate two completely new scenes but serves as a transition between (to use his terms) an exposition (background preparation) and a specific crisis in a new scene. In the previous section there is a sense of preparation for the one following, but there is still a discontinuity in tine marked in both cases.

[20] Cf. e.g. GK, para.116r and Ogden, VT 21, pp.451-469.

[21] See I. L. Seeligmann, 'Hebräische Erzählung und biblische Geschichtsschreibung', TZ 18 (1962), pp.310-311.

[22] E.g. Gn. 15:1, 22:1, 20; 25:26; 48,1. For other references see S. Mandelkern, Concordance on the Bible (New York: Schulsinger Brothers, [2]1955), pp.35 sub , pp. 35-37; cf. Irvin, Mytharion, p.27.

[23] Gn. 4:25 referring to verse 1; 8:12 to verse 12 to verse 6 and verse 12 to verse 11, 9:11, 15 to 7:21; 29:35 to verse 32; passim in chapter 30; 35:9 to verse 1.

[24] Gn. 3:22 (taking from the tree of life) referring to verse 6; 4:26 (son born to Seth) to verses 17ff.; 10:21 (son born to Shem) to verse 2(?); 27:31 (Esau makes tasty stew) to verse 14, 38 (another blessing for Esau) to verse 27, cf. verse 35; 32:20 (report of Jacob's arrival) to verse 18; 35:17 (another son born to Rachel) to 30:23.

[25] Cf. also Grayson, Chronicles 2-7, probably 8, 9-12, 15 and 16. 1:6-8 is a distinct section concerning an action continuing through several years. While it does not start with a specific year number as do the other sections, it does commence with a marginal time-reference ('at the time of '). The line was omitted between iii 18-19 in BM 92502 but is included in BM 75976; see Grayson, Chronicles, p.80.

[26] As noted by Grayson (Chronicles, p.84, commentary on iv 19-22) the order of months in iv 19-22 seems to be confused, Shubria being sacked in (month 10), its booty taken to Uruk in Kislev (month 9) and the king's wife dying in Adar (month 12). This could possibly be due to the time at which Esarhaddon became king. He had been crown-prince from Nisan (month 1), 681 BC and his father, Sennacherib,was murdered in (month 10; cf. RLA I, p.201), 681 BC. Although he did not gain actual control of the throne until Adar (month 12), 680 BC, he could have counted his kingship from the month his father died, which would place Tebet before Kislev in the first full year of his reign. Cf. also Grayson, Chronicles 17 ii 12-14 and iii 10-11 and the notes on them for other dischronological accounts. The former might be explained as was the example just noted, but the latter specifically mentions year 19 before year 16.

[27] See N. Richter, Exegese als Literaturwissenschaft (Göttingen: Vandenhoeck & Ruprecht, 1971), p.86; Schicklberger, TZ 34, p.70.

[28] Oxford English Dictionary (Oxford: Clarendon Press, 1933), VII, p.453 sub 'paragraph' 2, 'The words of a distinct speaker'.

[29] Andersen, Sentence, pp.77-78; cf. also F. I. Andersen, The Hebrew Verbless Clause in the Pentateuch (Nashville: Abingdon, 1970), p.35 and S. R. Driver, A Treatise on the Use of the Tenses in Hebrew (Oxford, Clarendon Press, 1892), p.201.

[30] See Andersen, Sentence, p.87; also pp.79-80 for further examples.

[31] Ibid pp.80-82 where examples are given.

[32] E.g. Grayson, Chronicles 23, 24. These two are both eclectic (see ibid. p.60) and, although they start each section by a marginal time reference, could have only one per reign so that subject and time changes are the same.

[33] R. R. Wilson, Genealogy and History in the Biblical World (New Haven/London, Yale UP, 1977, cited as Genealogy). See pp.207-215 for a detailed bibliography.

[34] Ibid., p.199.

[35] H. G. M. Williamson, Israel in the Book of Chronicles (London, Cambridge UP, 1977; cited as Israel), pp.62-63; cf. also M. D. Johnson, The Purpose of Biblical Genealogies (London, Cambridge UP, 1969), p.73.

[36] Cf. Williamson, Israel, p.63; A. Guillaume, 'Paranomasia in the Old Testament', JSS 9 (1964), p.283, n.2.

[37] Cf. also Gn. 36:15-19, 20-28, 29-30, 40-44 for other examples; Wilson, Genealogy, p.167.

[38] Coates, Canaan, p.9, cites Gn. 4:16, 13:18, 19:30; 20:1, 21:20, 21; 22:19; 26:6, 17; 50:22.

[39] Other such transitions are Gn. 13:12; 24:62; 25:11b; 29:14 (with a time reference); 47:27.

[40] gar - Gn. 21:34; - 21:33; cf. - 25:18.

[41] Seeligmann, TZ 18, pp.307-310; see J. Wilcoxen, 'Narrative', Old Testament Form Criticism, ed. J. N. Hayes (San Antonio: Trinity UP, 1974), p.91, and G. M. Coats, Rebellion in the Wilderness (Nashville; Abingdon, 1968), pp.47-48 and passim; cf. C. Westermann, Genesis 12-50 (Darmstadt; wissenschaftlische Buchgesellschaft. 1977), pp.49-51.

[42] See Westermann, Genesis 12-50, p.159.

[43] E.g. 12:8; 36:8; 38:11; 50:7-10 (/'ataq)

[44] E.g. 31:25 (taqa/nasag); 33:18 (); 35:21 (/); 38:1 /).

[45] Gn. 8:11; 22:8-9 (see verses 3, 6); 28:10-11; 35:5-6; 43:15; 44:13-14; 45:24-25; 46:1, 5, 50:7-10; cf. 37:14; 46:28.

[46] Only departure - Gn. 8:18-19; 14:8, 17; 34:1, 6; 37:12; 41:45-46; 47:10: only arrival - Gn. 7:7, 13; 14:5,13; 19:1, 23; 26:32; 27:30; 33:18; 34:20; 35:27; 37:23; 41:57; 42:5, 29; 44:14; 47:1.

[47] Gn. 12:9; 33:17; 35:16; hlak - Gn. 13:3; 14:11-12; 18:16; 21:14; 24:10, 61; 26:1, 31; 28:5, 7; 29:1; 31:55 - 32:1; 42:26; yrad- Gn.11:5: 12:10; 39:1; - Gn. 13:1; 17:22; 26:23; 35:13; shb- Gn.33:16.

[48] See from different periods, for example Dt. 34:1 (); Ru.1:1 (hlak); 2 Chr.8:3 (hlak).

[49] T. Jacobsen, The Sumerian King List (Chicago: University of Chicago Press, 1939).

[50] Grayson, Chronicles, p.140:4, 7, 10; cf. also col. v in which divisions were made according to the different locales from which the rulers came, i.e. 8-bala; l2-bala é-[I]bazi; 15-bala [ela][ki]; see p.197.

[51] E.g. King, Annals, p.49:7-8; 52:35-39; 59-60:7-9; 75:67 (Tiglath-pileser I) and passim in other royal inscriptions.

[52] See S. E. McEvenue, The Narrative Style of the Priestly Writer (Rome; Pontifical Biblical Institute, 1971), pp.16, 17, 158-169.

[53] A marginal time reference indicating a discontinuity between one section and those before and after; cf. section 1.1. above.

[54] Cf. C. Westermann, Genesis I-XI (Neukirchen-Vluyn: Neukirchener Verlag, 1974) p.117; 5. W. Anderson, 'A Stylistic Study of the Priestly Creation Story', Canon and Authority, eds. G. W. Coats and B. O. Long (Philadelphia: Fortress Press, 1977), pp.151-154; McEvenue, Style, p.17.

[55] Ibid., pp.77-78.

[56] Ibid., p.155; cf. N. Lohfink, Die Landverheissung eis Eid (Stuttgart; Katholisches Bibelwerk, 1967), p.45.

[57] McEvenue, Style, pp. 158-159.

[58] Gn. 2:4; 5:1; 6:9: 10:1; 11:10, 27; 25:12, 19; 36:1, 9; 37:2.

[59] The arguments are presented in Appendix C of the author's PhD thesis 'Some Scribal Techniques in Ancient Israel with Other Semitic Parallels' to be submitted to the University of London.

[60] O. Eissfeldt, 'Biblos geneseos', Kleine Schriften III (Tübingen: J. C. B. Mohr, 1966), pp.461-462. Cf. J. Scharbert, 'Die Sinn der Toledot-Formel in Priesterschrift', Wort-Gebot-Glaube, ed. H. J. Stoebe (Zürich: Zwingli Verlag, 1970), p.45, who notes that the formula occurs at different turning points in history.

[61] Time change: Gn. 2:4; 5:1; subject change:; Gn. 6:9; 10:1; 11:10, 27; 25:19; 35:1; time and venue, Gn. 25:12; subject and venue; Gn. 36:9; subject, time and venue:Gn. 37:2.

[62] Gn. 9:19[2]; 10:5, 20, 31[2]; 36:10 and passim.

[63] Gn. 36 passim. See Schicklberger, TZ 34, p.70.

[64] Cf. also the summaries in Gn. 8:1b; 23:20; 30:43; 31:21.

[65] See J. Lundbom, Jeremiah: A Study in Ancient Hebrew Poetry (Missoula; Scholars Press, 1975), p.16.

[66] The inclusio is recognized by Muilenburg in JBL 88, p.9.

[67] Coats, Canaan, p.9.

[68] Sentence, p.81; see also McEvenue, Style, p.38, concerning the 'echo'.

[69] Ibid., p.38.

[70] Also called variously 'concentric inclusion', 'concentric structure' or 'complex inclusion', the structure procedures through a series and then returns through the same series in the form ABCC'B'A; cf. ibid., pp.29, 157-158.

[71] See McEvenue's discussion of the chiasm of Gn. 17 in ibid.

[72] Cf. R. E. Longacre, 'The Discourse structure of the Flood Narrative', SBL 1976 Seminar Papers (Missoula: Scholars Press, 1976), pp.236, 241.

[73] Cf. Seeligmann, TZ 18 (1962), p.315.

[74] E.g. Gn. 20:1 referring to 18:1; 32:29 to verse 23. See Schicklberger, TZ 34, p.79.

[75] E.g. Gn. 1:28 to verse 30; 15:2 to verse 4; 20:7 to verse 14; 22:2 to verse 9; 24:2 to verse 9; verse 3 to verses 15ff., verse 4 to verse 10; 26:4-5 to verse 22; 27:34 to verses 39ff., 28:2 to verse 7; 29:15 to verse 18; 31:24 to verse 29.

[76] E.g. Gn. 3:17 to verse 3; 7:9, 16 to verse 2; 12:4 to verse 1; 17:23 to verses 10-11; 27:19 to verses 3-4. See Richer, Exegese, pp.85-88.

[77] Other examples are Gn. 24:16-25, 45-47 (Rebekkah); 32:10-23 (Jacob), 25-28 (Jacob's adversary). Cf. Richter, Exegese, pp. 85-88; Schicklberger, TZ 34 p.78.

[78] Richter, Exegese, pp.85-88.

[79] For a note of this division, see Andersen, Sentence, p.80. Other examples uniting different sections are Gn. 22:1 to chapter 21; 24:22 to verse 16; 26:26 to verse 18; 29:9 to verse 3; 32:23 to verse 10; 33:3-4 to verse 1.

[80] E.A. Speiser, Genesis (Garden City: Doubleday, 1964), p.122; G. von Rad, Genesis (London: SCM, 1961), p.202. other examples uniting two sources are Gn. 26:15 (S. R. Driver, Genesis, London: Methuen, 1926, p.252) to verse 13(J): 28:13(J) to 12(E): verse 19(J) to verse 11(E); 30:4(J) to verse 2(E); verse 20(J); Speiser, Genesis, p.229) to verse 19(E); 32:9(J); von Rad, Genesis, p.313) to verse 2(E).

[81] Cf. GK, para.126d; P. Jouon, Grammaire de l'hebreu biblique (Rome: Institut Biblique Pontifical, 1947), para.137, 3; Schicklberger, TZ 34, pp.80-81.

[82] Gn.1:3 ('light') is resumed by 1:4, 5, 18 (GK, para.126d); 22:2 ('one of the mountains') by 22:3, 4, 9 ('the place'); 28:1 ('stone') by 22:22.

[83] E.g. Gn. 1:26 ('man'; P) is presupposed by 1:27(P), 2:7, 8, 15, 16, 19, 20, 21, 22(J); 2:8 ('garden'; J) by 2:9,10; 3, 1, 8(J), 2:22 ('woman'; J) by 3:1, 4, 6, 13(J); 6:14 ('ark'; P) by 6:16, 18(P); 7:1, 7, 9(J), 13, 17(P) and passim; 21:2 ('son'; J) by 2:8 ('child', E); 24:2 ('servant') by 24:5, 10, 17, 22; 24:15 ('Rebekkah'; cf. verse 14: 'girl') by 24:28 ('the girl'); 25:24 ('twins'; P) by 25:27 ('the boys'; J); 12:10 ('famine') by 26:1; 29,1 ('land of the Easterners') by 29:22 ('the place'); 29:2 ('well') by 29:3, 10; 31:19 ('household gods') by 31:34; 33:2 ('women and children') by 33:5.

© 1980 A.R. Millard & D.J. Wiseman, reproduced by permission. Prepared for the web by Robert I. Bradshaw, January 2004. Please report any typographic errors.