Copyright © 1994 by Westminster Theological Seminary, cited with permission.
THE DIVISION AND ORDER OF THE PSALMS
R. DEAN ANDERSON JR.
RESEARCH into the psalter was revolutionized early in this century by
the work of H. Gunkel and his introduction of Gattungsforschung. This
work was subsequently developed by many, notably S. Mowinckel, and the
task and modern classification of Gattungen continues. Yet for the question
at hand, this modern research has not been all that helpful. D. J. A. Clines,
in his summary of research since 1955, for instance, states that "as yet a real
connection between psalms bearing the same title has still to be discovered."1
The various modern schemes of classification do not account for the present
arrangement of the psalter, nor for the psalter's own indications of arrange-
ment by its superscripts.2 Therefore this essay has concentrated upon what
I have considered to be basic issues. I shall consider first the basic textual
evidence for the psalter together with related issues arising from this evi-
dence. Thereupon follows a discussion of the basic division of the psalter
into five books. A discussion of superscripts and postscripts and their rela-
tion to the arrangement and order of the psalms ensues. Finally I go briefly
through the five books themselves attempting to show what may be said of
the arrangement of the psalms in each.3
II. Textual Evidence
The Masoretic psalter, as is commonly known, is divided into five books.
This division will be discussed below. The division of individual psalms in
the Masoretic tradition is not as uniform as one might suspect. There are
several individual cases in books 1-3 where mss show varying traditions of
combination of psalms. In books 4-5, however, many mss combine not a few
psalms in various ways. The most important of these combinations will be
discussed in more detail below.
1 D. J. A. Clines, "Psalm Research Since 1955: II. The Literary Genres," TynBul 20
2 Cf. G. H. Wilson, The Editing of the Hebrew Psalter (SBLDS
Press, 1985) 161-62.
3 Given the sheer volume of material written on the psalms, and the number of psalms
themselves involved, any conclusions reached in this essay should be regarded as tentative.
The evidence from
ought not to be overlooked in a study of our Masoretic psalter. This evi-
dence has engendered much debate, particularly over the question of how
the material is to be related to the MT. G. H. Wilson goes over this material
and its discussion in quite some detail, in particular considering the debate
between J. A. Sanders et al. and P. W. Skehan et al. over 11QPsa and its
of mss forming a reasonably sized collection of
merits attention.4 It is the more interesting because of its use of many
psalms in common with MT (from books 4 and 5), both in a different order
and in conjunction with psalms not known from
on palaeographical grounds to the first half of the first century AD.5
It is impossible in a paper of this size to go into any significant detail on
this matter, yet we ought to note the parameters of the debate and how it
affects our view of
sidered as a canonical and therefore authoritative, open ended canon of
psalms. He argues further that it precedes the completion of the MT psalter
as canon, forming an important step in that process. Thus, Sanders places
the completion of the MT psalter as canon at the end of the first century
Skehan has opposed this view arguing the reverse, i.e. that 11QPsa is
merely a liturgical collection with no real authority at all and no bearing
on the MT psalter as canon, which at least in its first four books was
complete by the fourth century BC, and the final section not much later. He
argues that 11QPsa is textually dependent on
shown that Skehan's criticisms go too far and cannot be sustained.7 There
is no real evidence that 11QPsa was dependent on MT. Both may well have
been dependent on a common tradition of psalm materials. Furthermore
both the MT psalter and 11QPsa seem to function as liturgical collections.
(It is important to note at this point that the discussion essentially concerns
4 See J. A. Sanders, The Psalms Scroll of
Clarendon, 1965); id., "Cave 11 Surprises and the Question of Canon," McCormick Quarterly 21
(1968) 284-98; P. W. Skehan, "The Qumran Manuscripts and Textual Criticism," in Volume du
Congris Strasbourg 1956 (VTSup
4; Leiden: Brill, 1957) 148-58. All other
are more fragmented.
making judgments based on fragmentary texts" (Editing, 67).
5 Sanders, The Psalms Scroll, 9.
6 For a critique of the notion of a late first-century council at Jamnia that determined the
extent of the canon, see R. C. Newman, "The Council of Jamnia and the Old Testament
Canon," WTJ 38 (1976) 319-49.
Old Testament Canon of the,
Early Judaism [
THE DIVISION AND ORDER OF THE PSALMS 221
books 4 and 5 of the psalter.8 The
whelmingly supports the arrangement of MT books 1-3 as we know them.
11QPsa begins at Psalm 101, containing no earlier psalm than this except
edition based on MT,9 seems unsure whether to see a parallel development
of MT books 4 and 5 and 11QPsa, or whether to suppose (with Sanders)
mately led to the MT arrangement of books 4 and 5.10 I propose that a
hypothesis of parallel development is more likely to be correct. There is in
the first place evidence that the Masoretic psalter even in books 4 and 5 was
extant at least contemporaneously with the community
evidence comes in the form of both Josephus and the LXX. First, Josephus'
(c. AD 37-110) earliest work. Against Apion, mentions the following in a
discussion of the (for him completed) canon: ai[ de> poipai> te<ssarej [sc. tw?n
bibli<wn ] u!mnouj ei]j to>n qeo>n kai> toi?j a]nqrw<poij u[poqh<kaj tou? bi<ou
perie<xousin.11 Secondly, although the date of the LXX psalter may be
obscure, it must at least have been completed well before the first century
when it was known throughout
by the NT).12
What then are we to say of 11QPsa? That it was regarded as containing
all canonical psalms seems highly probable in view of the Davidic prose
section. This section seems to ascribe Davidic authority (if not authorship)
to the entire collection. David as the inspired psalmist par excellence has a
central place within this collection.13 The so-called apocryphal psalms
8 The only evidence against MT books 1-3 are two mss (4QPsa, frgs. c, d; 4QPsq, col. I)
containing Psalms 31 and 33 (thus excluding Psalm 32), and one ms (4QPsa, frg. g) containing
Psalms 38 and 71.
9 In addition to the arguments above, this view is based on two pieces of evidence: (1)
more than one copy of this "psalter" has been found (cf. 11QPsb), and (2) the prose section near
the end of the MS gives David's last words (as the inspired and prophetic psalmist) and a tally
of David's compositions. Both these factors, in
functioned as canon (cf. Editing, 66). The second factor seems to me stronger than the first.
is not possible to demonstrate the existence of a
single consistent Psalter tradition at
Sanders, in his review of
criticism of his own position.
11 "The remaining four books contain hymns to God and precepts for the conduct of
human life" (Ap. 1.40; LCL translation).
12 One would suspect that the psalter, being so crucial liturgically, would have been trans-
lated soon after the Pentateuch. H. B. Swete, in addition to the evidence of NT quotations,
shows usage from early non-Christian Hellenists (An Introduction to the Old Testament in
Greek [rev. R. E. Ottley;
NT speaks of the psalms as a "book" (singular) in Luke 20:42 and Acts 1:20. This seems to
imply a finished canonical product.
13 We should be careful
here (contra Sanders; see
the implications of the prose section. That it does not imply Davidic authorship for the whole
contained here apparently bear no mark of having been written by the
shown that these psalms exhibit a fully biblical character.14 This fact tends
to lend support to the apparent
psalms of Davidic (temple?) origin. Could we perhaps regard these extra-
biblical psalms as indeed examples of more Davidic and prophetic com-
positions than those known to us, dating back to the times of David and the
prophetic schools following him that provided compositions for the temple
liturgy? Our present MT psalter seems to be no more than a collection of
what must once have been a rather massive supply of temple psalms.
11QPsa itself testifies to a much larger corpus of material when it notes that
David's compositions alone totalled 4,050.15 The Qumran sect thus, when
of psalms from a temple depository of prophetic liturgical material.16
The LXX is basically a translation from a proto-Masoretic type copy of
the five books of psalms. Our main interest in it concerns its witness to the
division of psalms and to superscripts. With respect to the division of
psalms, the LXX also has 150 psalms; however, they are divided somewhat
collection is clear from the inclusion of Psalm 127 with the same superscript as the MT
denoting Solomon as author (cf. MT Ps 72:20). Furthermore, there is no evidence that super-
scripts were deliberately altered or edited at
authorship of any given psalm material. Thus the superscripts in 11QPsa often fail to indicate
an author. In this ms and
ingly agree with
The differences consist in both additional material (compared to MT) and less material. Thus
it cannot be fairly stated that
is rather weak when he suggests, on the basis of two instances where 11 QPsa adds dvdl, as over
against one instance where it omits the same,
130)! In fact, a study of
and MT superscripts and postscripts, shows that the
number of times where
scripts or postscripts contain less material than MT far outweigh the cases where they contain
more. In verifiable cases,
interest is the fact that, in total,
lacking such in MT and twice contains no reference to Davidic authorship where MT has it.
my judgment, the differences on this score between MT and
to suggest that neither text should be designated expansive, but that the witnesses of both are
valuable attestations to possible original documentation.
15 It should be noted that 11 QPsa in no way claims that such a number of psalms and
songs from David were extant at the time of its compilation. Of interest, however, is also the
ninth-century AD report of a find of biblical and extrabiblical Hebraic writings in a rock-cave
lation see O. Braun, "Ein Brief des Katholikos Timotheos," Oriens Christianus (1901) 299-313.
16 In this respect we ought to remember the necessity of distinguishing between a
sectarian community and the biblical texts that such a community may use. Admitting that
one's judgment on the quality of the text that they used. In this case there is no evidence to
suggest that the community deliberately altered any biblical text material that they brought with
THE DIVISION AND ORDER OF THE PSALMS 223
differently. Two pairs of psalms in MT are joined in LXX (9 and 10, 114
and 115), and two psalms in MT are split in two by the LXX (116 and 147).
This evidence will be weighed below.
An analysis of LXX superscripts shows that, apart from a section of book
4 and book 5 of the psalter respectively, there is very little variation from
MT. Where additional material occurs it seems to refer mostly to the litur-
gical occasions when such psalms were sung (e.g., Psalms 24, 29, 38, 48, 93,
94—MT numbering). Such designations are by their own nature secondary
to the original psalm settings, yet that is not to say that the LXX translators
merely added them.17 In terms of the historical superscripts, there seems to
be only one case of secondary dependence on a historical book.18 There is
no evidence that the translators actually added anything to the text before
them,19 and of course, as is well known, superscripted liturgical information
is also present in MT (e.g., Psalm 92 and the songs of ascents, Psalms 120-
134). Occasionally the LXX has genre designations where there is only an
author given in MT (e.g., Psalms 11, 14, 25), but there is no consistency in
this at all, and there are far more examples where the LXX together with
MT give the author alone.
It has been alleged that the LXX exhibits a trend in ascribing more
psalms to Davidic authorship, and thus it is said to enhance the Davidic
authority of the psalter.20 This assertion is at best too generalized. A broad
overview of the psalter reveals very little difference between MT and LXX,
except for one clear group of psalms (Psalms 91-99). If we pass over this
17 In that regard they are similar to the MT superscripts which are often also secondary
to the original psalm settings.
18 The LXX superscript of Psalm 96 reads: o!te o[ oi#koj &]kodomei?to meta> th>n
ai]xmalwsi<an: &]dh> t&? Dauid ("when the house was built after the captivity; a song of
David"). This is sometimes taken to indicate that the translator knew that his ascription to David
was incorrect, since he also made the psalm postexilic by referring to the building of the second
temple; see F Delitzsch, Psalms, in Commentary on the Old Testament in Ten Volumes (repr.
Commentary on the Psalms (ICC; Edinburgh: T. & T. Clark, 1906) 2.299. However, this seems
pretty unlikely. It may be that a Davidic psalm was used at that time and so attained the added
historical superscript. However, it seems to me more probable that the reference here is to the
building of the tent for the ark after David
retrieved and brought it to
captured by the Philistines and so taken into captivity (1 Sam 4-6). This incident is referred to in
Ps 78:61 where LXX also translates ai]xmalwsi<an. Note that vzf here refers to the ark (cf.
Ps 132:8). The superscript probably thus derives from 1 Chr 16 where Psalm 96 is supposedly
utilized as part of the song of praise recorded by the Chronicler at this event. The historical
part of the superscript is therefore probably dependent on Chronicles. MT has a similar case
of superscript dependence on a historical book in Psalm 18, derived from 2 Sam 22:1 (cf.
B. S. Childs, "Psalm Titles and Midrashic Exegesis," JSS 16  139).
19 An exception to this may be Psalm 31 (LXX 30), where e]ksta<sewj seems to be
derived from the translation itself (LXX v. 23).
20 E.g., C. Th. Niemeyer,
Het problem van de rangschikking
der Psalmen (
Luctor et Emergo, 1950) 149, who includes the MT, 1 Chronicles, and NT in this trend; A.
Pietersma, "David in the Greek Psalms," VT 30 (1980) 213, who includes the MT in this trend;
special section of book 4 for a moment (to be discussed below), additional
ascriptions to David occur only five times. Of
attest Davidic authorship. It should also be noted that twice the LXX lacks
designation of Davidic authorship where MT has it (Psalms 122 and 124)!
Few would argue that MT is expansive here!21
Overall, in cases where LXX contains superscript material not found in
MT, it seems clear that the translators were nevertheless reading a Hebrew
Vorlage. This is shown by the fact that in these so-called additions, equally
difficult Greek (relying on semitic constructions) is utilized (e.g., t&? for l
of authorship; improbable nouns in genitive constructions wrongly interpret-
ing supposed Hebrew bound constructions, such as ai#noj &]dh?j). That is
not to say, however, that this Hebrew Vorlage is always correct. Its textual
value must be assessed in each case, comparing evidence from MT and
Psalm 99 (LXX) is highly likely on internal evidence (see comments below),
whilst on the other hand the LXX ascription of Davidic authorship to
Psalms 43 and 137 is extremely unlikely.22
It is interesting to note the "additions" to Psalms 91 and 93-99. Here,
where the MT for the most part provides no superscripts at all, the LXX
consistently has rather full designations. Similarly, the superscripts to
Psalms 143-48 (including a variant division) in book 5 are noteworthy.
What makes this phenomenon so interesting is that it occurs precisely in the
last two books of the psalter.
It is also these last two books that in
were organizationally most variant from
ports the theory that books 1-3 of the psalter were very early stabilized, and
that mss of books 4 and 5 continued to exhibit fluctuation to a later date,
probably indicating that they were put together at some later time.
21 In this respect Pietersma quite incorrectly alleges that "all that is in the MT is in the
LXX" ("David," 214). He seems to acknowledge the two cases in point on p. 217 n. 11.
22 Pietersma ("David") argues that the extent of additional material in the LXX super-
scripts actually going back to a Hebrew Vorlage may be less than the text of A. Rahlfs indicates
(ed. Psalmi cum Odis [
parts of the LXX text tradition show a tendency to increase Davidic titles. Nevertheless, I am
not convinced that those superscripts overwhelmingly attested in the mss tradition are later
additions. In general, Pietersma's argumentation is not always based on much evidence. For
instance, he argues that since o!te is only found in superscripts omitted in MT, it is therefore
probably secondary (p. 221). However, the total number of historical superscripts (where o!te
might be expected to be found) is only 13 in the MT tradition anyway. This is far too small
a witness to function effectively as a control on vocabulary! Cautions here not sufficiently
noted by Pietersma are (1)
the presence of variant Hebrew
that even in Pietersma's view at least in one case the LXX omits reference to David where MT
has it (Psalm 124 MT), and (3) the fact that even Pietersma admits that there are cases where
a different Hebrew Vorlage is likely (e.g., Psalm 104 MT), and these could lead us to suspect
THE DIVISION AND ORDER OF THE PSALMS 225
III. Division into Books
As is well known, the Masoretic psalter is divided into five books. This
basic arrangement is also attested by the LXX. As noted above, however,
books 1-3 seem to have been well established much earlier than books 4
The division into five books is clearly supported by various indications in
the psalter itself. In the first place each book seems to end with an appro-
priate doxology (cf. Ps 41:14; 72:18-19; 89:53; 106:48). It has often been
suggested that Psalm 150, or alternatively 146-50, should be seen as the
doxology for book 5. This will be discussed below. Niemeyer and Wilson are
probably correct when they argue that these doxologies should be seen as
integral parts of the psalms they are attached to, and not as editorial
additions.23 Their appearance at the ends of the respective books, however,
is highly suggestive of deliberate placement.
Another way that these books are distinguished is by the use of the
divine name. Book 1 clearly prefers the name Yahweh, utilizing it 273
times (as opposed to Elohim, 15 times). Book 2, on the other hand, prefers
Elohim (164 times, as opposed to Yahweh, 30 times). Book 3 is mixed. The
Asaph psalms (73-83) clearly prefer Elohim; however, the rest of the book
prefers Yahweh. Books 4 and 5 both prefer Yahweh (book 4 contains no use
of Elohim, and book 5 employs it only 7 times, as compared to 236 uses
One final indication of the division into books has been argued by Wil-
son.24 He notes that the superscripts play an important role in this division.
Whilst no part of the superscripts may be seen as the primary editing
principle, there are certain important factors to note. First, each book is
clearly marked off from the preceding by a change of authorship. The only
exception here is the transition between books 4 and 5, which is more fluid.
He also notes that the transition between author groupings within any given
book is softened by overlapping genre titles. For example, the psalms of
Korah end at Psalm 49. Psalm 50 is Asaphite, and Psalms 51-65 are Davidic.
The fact that Psalms 47 through 51 are all marked mizmor softens the
transition. Compare also Psalms 65 through 68. Equally important is that
such overlapping genres do not occur between books.
In conclusion, then, we may justly infer that the division into five books
has not been imposed upon the psalter, but is inherent to its formation.
23 This is due both to the way they fit each particular psalm in question and to the non-
formal agreement between the doxologies themselves. See Niemeyer, Het probleem, 72-78.
Niemeyer goes on to reject the view that the doxologies function to close the individual books.
He notes that "geen enkele bundel of verzameling in het O.T. met een doxologie werd
afgesloten" (p. 76). While this may be, it ought to be noted that there is no other collection
of songs or psalms in the OT at all! See also
IV. Superscripts and Postscripts
The superscripts and postscripts to the psalms can lay claim to a fair
degree of antiquity. Not only are they evidenced in
and Targums, but it ought also to be noted that even by the time of the
LXX translation (second or third century BC?) the technical terms con-
tained there were so antiquated and obscure that the translators had a fair
degree of trouble interpreting them. This is true also for the Targums.
Furthermore, we find similar super/postscripts in other parts of Scripture
(cf. Hab 3:1, 19b; Isa 38:9). There thus seems to be no reason not to take
the super/postscripts seriously.
Given that attribution of authorship seems to play an important role in
the division of the psalter, it behooves us to investigate briefly the signifi-
cance of l before personal names in the superscripts.25 The most obvious
denotation of lamed before a personal name seems to be that of authorship
(commonly called lamed auctoris; cf. Ges-K 129c). A fairly clear example of
this in a superscript outside the psalter is Hab 3:1. Within the psalter itself,
it seems highly probable that in those cases where dvdl is followed up by a
historical superscript placing the psalm in a part of David's life setting,
authorship is meant. Certainly in Ps 18:1 this is the case (cf. 2 Sam 22:1).
Furthermore, the NT seems to have taken the lamed in this way; note
Matt 22:41-45; Mark 12:35-37; Luke 20:41-44 (Psalm 110); Acts 1:16-17
(Psalm 41); 2:25-34 (Psalms 16 and 110); Rom 4:6-8 (Psalm 32); Rom 11:9f
(Psalm 69); Heb 4:7 (Psalm 95 following LXX).26
Despite all this, it is equally clear that lamed plus personal name does not
always indicate authorship. The phrase Nvtydyl in Ps 39:1, for instance, must
indicate something other than authorship, which seems to be indicated by
the ensuing dvdl. J. Ridderbos argues with some plausibility that the phrase
here specifies the choir director more closely.27 The phrase Hrq-ynbl is also
not so clear. The fact that no other author indication is given in these
superscripts would seem to suggest authorship. However, Ridderbos, fol-
lowing Weiser, on the strength that here no personal name is given, suggests
"belonging to."28 This meaning seems to me at least possible. With respect
to the Asaph psalms, Ridderbos rejects lamed auctoris, but only because he
finds it difficult to consider some of these psalms as dating from the time of
the Davidic Asaph.29 This argument is weak. If the authorship of the Davidic
Asaph is to be rejected, and the possibility of a later Asaph seems improb-
able, then the respective titles are probably spurious. The dating of several
Asaph psalms will be discussed below In conclusion, apart from such cases
25 My discussion here is deeply indebted to J. Ridderbos, De Psalmen (Commentaar op
het Oude Testament; 2 vols.; Kampen: Kok, 1955-58) 1.367-381.
26 For these examples see ibid., 374.
27 Ibid., 376.
28 Ibid., 376-77.
29 Ibid., 377-78.
THE DIVISION AND ORDER OF THE PSALMS 227
as Nvtydyl or Hrq-ynbl, I take lamed plus personal name in the superscripts
to refer to authorship.
The last problem discussed here, as far as the superscripts are concerned,
relates to the so-called historical superscripts. Many of the psalms bearing
these seem to lack any concrete detailed historical reference confirming the
superscripts themselves. This problem is, however, more acute in some
instances than in others. Psalm 51, for example, is widely acknowledged as
portraying the penitent spirit of one in a position such as David found
himself.30 Psalm 34, on the other hand, has been judged more problematic.
N. H. Ridderbos has suggested that in such a case the psalmist may well
have written or used the psalm in question on the occasion indicated, but
modeled it according to a well known and generalized form.31 This view,
of course, also makes the psalms themselves particularly suitable for preser-
vation and continued use by the covenant people throughout their genera-
tions and different circumstances of need and praise.
B. S. Childs, however, argues that the historical titles in no way depend
on tradition or historical memory. In his view they are the result of mid-
rashic exegesis that was taking place in the few centuries before Christ.
Childs bases his working hypothesis here upon evidence related to Psalm
Relying upon Sanders' comparison of LXX Psalm 151 and
Psalm 151, he suggests that the LXX superscript to Psalm 151 must be a
later addition that fits only the expansive amalgam of two earlier psalms
that form LXX Psalm 151.32 He continues: "This evidence is significant not
only in showing the lack of an independent historical tradition behind the
expansion of the LXX titles, but also in fixing an approximate date at
which time Psalm titles were being formed on the basis of internal develop-
ment alone."33 Thus, Childs clearly moves from an argument based on the
superscript of LXX Psalm 151 to an extrapolation about all the LXX
superscripts, finally bringing forward a hypothesis on historical psalm super-
scripts in general. This argumentation is, however, rather weak. In the first
place, Sanders cogently argues that LXX Psalm 151 is a translation of a
Hebrew amalgam of Qumran Psalms 151A and 151B.34 Thus, the LXX
translators probably should not be considered guilty of supplying their own
titles based on midrashic exegesis. Secondly, we may ask whether the ac-
tivity behind a clearly apocryphal psalm composition can be said to reflect
similar activity behind the canonical psalms. This question receives especial
significance in view of the fact that the LXX itself places Psalm 151 outside
30 See for example B. S. Childs, "Psalm Titles," 145.
31 N. H. Ridderbos, Psalmen en cultus (Kampen: Kok, 1950) 20-22. It is also interesting
to note that E. J. Young (An Introduction to the Old Testament [2d ed.;
1949] 301) uses this phenomenon to argue that the psalm titles themselves must be authentic.
How else can their connection to otherwise unrelated psalms be adequately explained?
32 Sanders, The Psalms Scroll, 56ff.
33 Childs, "Psalm Titles," 143.
34 Sanders, The Psalms Scroll, 56ff.
the canon.35 Finally, can the supposed literary activity of the LXX be
automatically predicated of the MT? Childs' view is thus found severely
wanting in its premises. Nevertheless,
conclusion that this midrashic work probably "stemmed from a pietistic
circle of Jews whose interest was particularly focused on the nurture of the
superscripts within the psalter is to loosen those psalms from their original
cultic context and give them an individual spiritual character that makes
them appropriate for exemplaric meditation. To me this seems no more
than speculation based on a wrong view of the nature of the superscripts
themselves. If the MT psalter really was intended to provide such a
"hermeneutical key" for individual meditation in this way, why is the
phenomenon of historical superscripts not more common and more orga-
nized? As it is, these superscripts are spread out over a good portion of the
psalter. In conclusion, there seems to be no real editorial purpose to either
the existence or the present distribution of the historical superscripts.
V. Books 1-3
1. Division of Individual Psalms
Before proceeding to a discussion of the final arrangement of the five
books of the psalter itself, we must first briefly examine the main cases put
forward for combination and separation of psalms contrary to their present
state in codex Leningradensis.
that we need to treat the cases in books 1-3 separately from those in books
4 and 5. His own examination of the candidates for combination here falls
under a discussion of untitled psalms.37 Having examined these and the
grounds for their combination with those psalms preceding them, he con-
cludes that the MT placement of untitled psalms in these books is so as to
deliberately connect them with the preceding psalm. He argues that in
each case valid reasons for their combination can be given. It remains for
us to test this for ourselves case by case.
Psalms 9 and 10 are frequently taken together as forming one acrostic.
Psalm 10 has no superscript and is joined with Psalm 9 in LXX, Vg, and
a few Hebrew Mss. Scholars are not unanimous on this union, however.38
35 LXX states: ou$toj o[ yalmo>j i]dio<grafoj ei]j Dauid dia> e@cwqen tou? a]riqmou?
("This psalm is a personal writing of David and outside the number").
id., "Reflections on the Modern Study of the Psalms," in Magnalia Dei: The Mighty Acts of God:
Essays on the Bible and Archaeology in Memory of G. Ernest Wright (ed. F M. Cross et al.;
Garden City, NY: Doubleday, 1976) 384.
37 Editing, 173ff.
38 See e.g., N. H. Ridderbos, De Psalmen (Korte verklaring der Heilige Schrift; Kampen:
Kok, 1962) 1.126-27, who follows F Delitzsch, Psalms 1.175-76.
THE DIVISION AND ORDER OF THE PSALMS 229
This is mainly because the acrostic no longer seems to be completely pre-
served. Whilst Psalm 9 probably gives us x to k (missing d), and Psalm 10
begins appropriately with l the following verses show no evidence of con-
tinuing the sequence until v. 6b (p), or possibly not till v. 8b (f) or even
v. 12, where it continues from q to t. It should be noted, however, that the
number of missing letters equals the numbers of verses lacking acrostic
beginnings. In my opinion the remaining evidence of an acrostic cannot be
discounted, and the original unity of the composition thus seems fairly
evident. Furthermore, there are several internal indications of unity be-
tween Psalms 9 and 10. Note the use of similar phrases, e.g., tvtfl hrcb
(9:10; 10:1), and the use of vz with similar imagery (9:16; 10:2), and compare
9:18-20 with 10:11-12.39 A distinction between Psalms 9 and 10 has been
observed in the content. Psalm 9 is characterized as a psalm of thanksgiving,
whilst Psalm 10 as a supplicatory psalm. Further, Psalm 9 is said to be
concerned with enemies from the outside, whereas Psalm 10 with enemies
from the inside.40 The first distinction does not seem to carry too much
weight in view of the inner verbal connections. The entreaty of Psalm 10
is clearly based on the confession of Psalm 9 (which is itself not devoid of
entreaty; cf. vv. 20-21).41 The nations are mentioned in both psalms as
enemies (cf. 9:18--21 and 10:16). It is understandable that it is in the suppli-
cation proper that a more detailed and expanded view of oppression is
given (i.e. allusion to enemies within).
Wilson, who accepts an originally unified composition, argues that the
break in the acrostic pattern at the beginning of Psalm 10 can only be
explained by assuming that the psalm was early broken into two distinct
compositions/readings.42 He may well be correct in this assumption,
though to infer a division in readings does not in my opinion necessarily
have to entail an early division into separate compositions. A tradition of
readings separating these psalms may well account for the separation of
them in many Masoretic Mss.
Psalms 32 and 33 are sometimes said to be a unity.43 Here there is indeed
some external evidence. A few Masoretic mss apparently join the two
psalms together. Furthermore, F. Delitzsch cites an old midrash that
reckons them together.44 Against this evidence, however, is that of both the
39 See further Delitzsch, Psalms 1.175.
40 Cf. ibid; N. H. R.idderbos, De Psalmen 1.126-27; and J. Ridderbos, De Psalmen 1.77f.
41 Note that De Wette classified Psalm 9 as supplicatory (see J. Ridderbos, De Psalmen 1.78).
43 See for example
Psalms 9/10; 32/33; 42/43; and 70/71 as single psalms. On pp. 174-76 he argues not for an
original unity but for a secondary unity based on "an early liturgical function." Although this
helps to give a clearer authorship pattern, the evidence for this early conjunction seems to me
44 It also joins Psalm 42 with 43, and 9 with 10. See Delitzsch, Psalms 2.54. Whilst this
evidence may be significant,
but also attributes the latter to David (i.e. it gives it a separate super-
script).45 Two Qumran fragments contain the order of Psalm 31 followed by
Psalm 33.46 Psalm 32 is thus not connected. 4QPsq also gives Psalm 33 a
superscript attributing authorship to David.47 Finally, there is the fact that
there seems to be no internal connection between the two psalms. The
evidence of the LXX and
are the few Masoretic Mss. It is fairly easy to see how, in copying a collection
of psalms predominantly separated by superscripts, two psalms not so sepa-
rated could accidentally (or even perhaps intentionally) be joined. This is
especially likely in view of the fact that the last verse of Psalm 32 and the
first verse of Psalm 33 are so similar.48
tain the separation of Psalms 32 and 33.
Psalms 42 and 43 are frequently classed as one psalm. In support of this
union are many Hebrew MSS and the fact of striking internal unity. Further-
more Psalm 43, being without any title, is rather out of place by itself in the
midst of a group of Korahite psalms.
Psalm 42/43 is clearly set in 3 strophes, each ending with the same
refrain, "How you are in despair, 0 my soul." Psalm 42 contains the first
two refrains, and Psalm 43 the third and final. This is, however, not the only
connection between Psalms 42 and 43, for even within the strophes them-
selves there are many inner verbal connections. This occurs both between
the first two strophes in our Psalm 42 and between those strophes and the
third one in what is our Psalm 43. Furthermore the setting is quite clearly
the same in both. Finally, it is clear that each strophe builds on the pre-
ceding, up to the third which forms the climax and conclusion of the psalm,
without which Psalm 42 is sadly stunted.49 The first two present the position
of the psalmist far away from the temple urging himself to place his trust
in God and asking him for deliverance. This plea for deliverance finally
comes in the third strophe (Psalm 43). Perhaps this distinction between the
first two and the final strophe (similar to Psalms 9 and 10) led to its division
in certain Masoretic Mss.
"problem" psalms without superscripts. This seems unlikely in view of (1) the fact that the
LXX is not consistent in this respect, as it joins Psalms 9 and 10 and thus shows no deliberate
methodology; (2) my argument above on LXX superscripts in general, i.e., that they were
based on Hebrew originals, and not fabricated by the translators; and (3) the support of
46 That is, 4QPsa frgs c and d, and 4QPsq.
47 It reads (column 1): rvmzm ryw dvdl.
48 This could well have been an ordering factor in the placing of Psalm 33 after 32 in the
49 Thus I do not concur with N. H. Ridderbos' suggestion that Psalm 43 may have been
added later to complete Psalm 42 (De Psalmen 2.13-14). The presence of a superscript for Psalm
43 in the LXX, however, does suggest that a separation between these two psalms occurred
THE DIVISION AND ORDER OF THE PSALMS 231
Psalm 71 also stands out in book 2 as being without a superscript. Here
too, many Masoretic mss join Psalm 71 with the preceding psalm. At first
sight this seems at least possible. As
connections between the two psalms (cf. 70:2b and 71:12b; 70:3 and 71:1,
13, 24).50 These internal connections, however, should not be overempha-
sized. They are not at all on the same level as those between Psalms 42 and
43. The LXX separates Psalm 71, preserving a superscript for it which
reads: T&? Dauid: ui[w?n Iwnadab kai> tw?n prw<twn ai]xmalwtisqe<ntwn.51
The intriguing reference to the sons of Jonadab (cf. Jeremiah 35) suggests
that the (Davidic) psalm was used by them and the first exiles sometime
after disaster struck in 587 BC. There seems to be no literary connection to
Jeremiah 35, and so this historical superscript probably relies on ancient
Here in 4QPsa (frg g) Psalm 71 follows on from Psalm 38. On balance, then,
the evidence strongly suggests that Psalm 71 ought to be treated as a psalm
independent of Psalm 70. Those Masoretic mss joining them probably did
so either by accident or to avoid an unseemly exception to the use of super-
scripts (see the argument above for Psalms 32 and 33).52 The internal
connection between the two psalms may explain why they were placed next
to each other in the psalter.
Having examined those psalms sometimes combined, it is my conclusion
that although in certain cases good arguments can be made for such com-
bination, there is no evidence that this should automatically be predicated
of all psalms in books 1-3 lacking a superscript. Rather, I have argued that
in certain cases the only reason some MT Mss combine untitled psalms with
the former is because of that very lack of a title. Such combinations are
therefore quite secondary. Furthermore, those psalms which ought to be
combined may well have been so unified before the compilation of our
psalter (cf. the evidence of LXX). In these cases MT separation may well
have occurred at some stage in the transmission of the text for a variety
of reasons (see suggestions above). I therefore
that the placement of untitled psalms in books 1-3 shows obvious edi-
2. Arrangement of Books 1-3
Given that Psalms 9 and 10 are a unity, and that Psalm 33 probably
ought to have a Davidic superscript (see above), we may quite appropri-
ately characterize book 1 as a Davidic book. That David was an important
51 "Of David; of the sons of Jonadab and the first men taken in captivity."
wrongly dismisses LXX evidence as expansive.
55 Editing, 181.
and prolific psalmist is well known (cf. 2 Sam 23:1-2; Amos 6:5). The only
psalms lacking direct Davidic ascription are Psalms 1 and 2. Given the
overall Davidic character of this book, these psalms probably also date from
his period, if not from his hand (cf. Acts 4:25-26). Psalm 1 has often been
seen as introducing the psalter. It at least forms an appropriate introduction
to book 1.
The book itself shows little evidence of deliberate internal ordered place-
ment. Just over half the psalms could be characterized as pleas to God for
deliverance, but these are not set apart as a special group.54 Superscript
phrases do not indicate any separate groups of psalms either. Occasionally,
elements in common between adjoining psalms can be noted which may
have been a factor in their placement together, but this phenomenon is by
no means great or consistent.55
Book 2 begins with a group of seven psalms belonging to the sons of
Korah. It seems best with J. Ridderbos and F Delitzsch to understand by
"the sons of Korah" the well-known group appointed over the service of
worship by David.56 The sons of Korah survived the judgment of Num-
bers 16 (see Num 26:11). They were gatekeepers from Moses to David
(1 Chr 9:19; 26:1-19). This seems to fit Ps 84:11. They had become re-
nowned in helping David (1 Chr 12:6), and were appointed over the service
of worship (1 Chr 6:31-32; 25). 2 Chr 20:19, in the time of Jehoshaphat,
seems to be the last time they are heard of. However, Psalm 87, with the
N. H. Ridderbos assumes that the sons of Korah did return after the exile
since "the sons of the gatekeepers" are mentioned as such (Ezek 2:42 et
al. ).57 This seems likely. Note that Heman the singer in the time of David
stemmed from Korah (through Abiasaph, 1 Chronicles 6). So it seems some
of the sons of Korah were gatekeepers, and some (through Heman?) were
singers (ergo 2 Chr 20:19). Asaph the singer stemmed from Gershon
(through Libni). The sons of Heman and Asaph together are identified as
singers (1 Chr 6:33-48; 2 Chr 5:12). The sons of Asaph are sometimes
identified as the singers (Ezra 2:41; 3:10-11; Neh 7:44; 11:17, 22; and per-
haps 12:46). But this designation does not necessarily seem to exclude the
Korahite line of Heman (2 Chr 35:15). Given that the sons of Korah are
never referred to by name in Ezra/Nehemia, but only as "the gatekeep-
ers," it may not be out of place to understand singers from their line being
subsumed under "the sons of Asaph." Is this because the name of Korah
still had bad connotations? Or perhaps the sons of Asaph gradually became
54 By my count (without going into detailed exegesis), there are 21 such pleas out of a
total of 40 psalms.
55 Niemeyer has criticized Delitzsch in this respect for going much too far (Het probleem,
126-28). Possible examples of common elements include Ps 3:6-7 and Ps 4:9; Ps 34:8 and Ps
35:5ff. (hvhy jxlm).
56 J. Ridderbos, De Psalmen 2.9; Delitzsch, Psalms 2.52.
57 De Psalmen 2.5.
THE DIVISION AND ORDER OF THE PSALMS 233
dominant in this respect, their name therefore becoming the title under
which the others were also assumed.
None of the seven psalms of Korah gathered here in book 2 seems to
demand a late date. In this respect Psalm 44 should be placed in the time
of the war with
Psalm 60 that this was a bloody war.58 Psalm 44 is closely related to
Psalm 60, the superscript of which also places it at this time (cf. Ps 44:10
with Ps 60:12). Evidently, several captives were taken and sold into slavery.
The mention of such sale (v. 13), and the fact that the psalmist claims
innocency for the people of God (vv. 18ff.), both argue against an exilic date
for this psalm.
The remaining psalms in this book also date from Davidic or Solomonic
times. Changes of authorship within the book are smoothed over by genre
groupings.59 Once again there are occasional indications of similarity be-
tween adjoining psalms suggesting a rationale for their placement, but the
phenomenon is not widespread.60
matic unity among Psalms 65-68 which helps bridge the gap in authorship
contained in Psalms 66 and 67.61 He suggests that whilst many psalms on
either side of this group are prayers for deliverance, these psalms are all
songs of praise. This point seems sustainable. We might also note that in
contrast to the psalms on either side, all four in question contain the title
ryw. In addition to this connection, however, it does not seem to me out of
place to suggest a further link between these psalms. Psalm 65 closes with
the pastures and valleys shouting for joy and singing. This is immediately
taken up in Psalm 66 where all the earth is urged to shout joyfully to God
and sing the glory of his name. Both psalms are also connected by the
mention of the paying (Mlw) of the vow (rdn). In the final section of this
psalm (vv. 16-.20), the psalmist implores the godly to come and hear him
tell of what God has done for his life. Thus, Psalm 67 gives the priestly
benediction and summons the peoples to praise God that his way may be
known. Psalm 68 follows this up with a recounting of what God has done.
The final psalm of the book ends with a suitable doxology, as noted
above. One very interesting item, however, is the line underneath this, "the
prayers of David, son of Jesse are ended" (Ps 72:20). Possibly tvlpt here
should be read with LXX (u!mnoi) as tvlht. Striking is the fact that this
occurs at the end of a psalm of Solomon! Evidently the line is to be taken
58 J. Ridderbos (De Psalmen 2.22-23) rejects the relevancy of 1 Kgs 11:15 by stating that
it is "zeer dubieus." This seems to me a very questionable argument, for the text can very well
be maintained (cf. NASB), not to mention the evidence of the superscript to Psalm 60 together
with the contents of that psalm.
59 That is, where adjacent psalms have different authors, they invariably belong to the
same designated genre.
60 Psalms 56 and 57, for instance, both begin Myhlx ynnH, whilst Psalms 57-59 all
contain tHwt-lx in their superscripts.
61 Editing, 190-91.
in a general sense.62 It may just refer to book 2, but likely refers to both
books 1 and 2. Given that Davidic psalms occur in all the remaining books,
it suggests that books 1 and 2 were completed at an earlier point in history
than books 3-5. There seems to be no reason to date their compilation any
later than the beginning of Solomon's reign. This would fit with the general
organizational activity around the building and equipping of the temple at
that time. However, it is also possible that they were compiled by the men
of Hezekiah together with the Asaph psalms (see below).
Books 1 and 2 as two collections, probably compiled around the same
time, supplement each other quite nicely. Book 1 is a collection of psalms
favoring the divine name Yahweh, whilst book 2 favors Elohim.63 Of interest
is the fact that there are several instances of textual relation between psalms
in book 1 and book 2. First, Psalm 14 of book 1 is textually related to
Psalm 53 of book 2. Here we see an example of how similar psalm material
could be reworked at a different time for a different situation. In accor-
dance with the preference of the respective books, each psalm addresses
God with the different titles. Further, Niemeyer has
shown that other
tual differences between the two psalms should not be attributed to textual
corruption, but may well indicate a reworking for a different occasion.64
Secondly, Psalm 70 is clearly related to Ps 40:14-18. Again there is no need
to harmonize the slight differences here. A portion of psalm material has
been used on a different occasion in a different setting. Finally, it is inter-
esting to note the relation between Ps 31:2-4a and Ps 71:1-3.65 It is note-
worthy that all these cases occur with psalms of Davidic authorship (note
that Psalm 71 has Davidic ascription from LXX). The only other case of
intertextual relation in the psalter is also Davidic. Psalm 108 is related to
Ps 57:8-12 and Ps 60:7-14. It is not surprising that a prolific author such
as David should rework his materials from time to time. It is significant that
no two examples of such reworking of the same textual material can be
found in any one book of the psalter.
Thus we turn to book 3. The first section of book 3 is made up of a
collection of psalms of Asaph. This collection is set off from the other psalms
62 In view of my comments above on lamed auctoris, however, and the relation of this
phrase to the superscripts, the 5 here ought still to be taken as marking authorship (contra
Niemeyer, Het probleem, 94).
63 N. H. Ridderbos (De Psalmen 2.13) argues that in many cases Yahweh has been
supplanted by Elohim in book 2. The reason is no longer known, but he suggests that maybe at a
certain time, in "een
bepaalde kring van
to me a little speculative. The distinction may be no more than stylistic; see also Niemeyer, Het
64 Het probleem, 82-85.
65 Psalm 86 is also sometimes said to be heavily dependent on other sources (see for
example Childs, "Reflections," 382). Most of the psalm, however, is readily explained merely by
allusion to common poetic phrases. The allusion to such a common, important, and well
known phrase as Exod 34:6 in v. 15 is not unexpected. Apart from that the only other probable
allusions seem to be from Ps 54:5 (v. 14), 25:1 (v. 4b), and Exod 15:11 (v. 8a).
THE DIVISION AND ORDER OF THE PSALMS 235
in this book in a number of ways. First, the rest of the psalms of book 3 show
no evidence of deliberate arrangement, by author or otherwise. Second, the
psalms of Asaph are all characterized by their preference for Elohim. The
ensuing psalms show a preference for Yahweh. Third, the psalms of Asaph
appear to bear a number of other characteristics in common. These are
enumerated by Niemeyer as follows: (1) the preference for lx and Nvylf;
(2) God called or summoned as judge; (3) interjections of God speaking;
(4) the use of the image of shepherd and sheep to signify God's relation to
his people; (5) mention of Joseph and the tribes stemming from him
(Ephraim and Manasseh); (6) allusions to ancient history; (7) a common
base of vocabulary.66
All this strongly suggests not only that these eleven psalms originally
formed a separate unit together, but also that they come from the hand of
the same psalmist. It is my contention that we should see here the hand of
the Davidic Asaph (see above under discussion of lamed auctoris). Asaph was
a chief among the singers appointed under David (1 Chr 15:16ff.; 16:4-7, 37;
25:1ff.). Asaph as such functioned as xybinA, (cf. 1 Chr 25:1-2) and in that
capacity appears to have been an important author of psalms (cf. 2 Chr 29:30).
That xbn in 1 Chr 25:1-2 must be seen in the technical prophetic sense is
clear from 2 Chr 20:14ff.
In objection to the Davidic Asaph's authorship of these eleven psalms,
many have turned to internal considerations manifested in Psalms 74, 79
and 83 in particular. Among the Asaphite psalms, 74 and 79 are often said
to be exilic, if not Maccabean.67 A careful examination of these psalms,
however, reveals that this may not necessarily be the case. Nowhere in these
psalms is there any mention of an exile. I submit that it is possible that these
psalms date around 926/925 BC when Shoshenq
We learn from the biblical account (1 Kgs 14:25-28; 2 Chr 12:1-12) that,
in the fifth year of Rehoboam,
horsemen, capturing the fortified cities Rehoboam had built (2 Chr 11:5-
He finally came as far as
bled themselves, was spared a little (Ffmk 2 Chr 12:7). Yet Shoshenq did
of events he then went on to attack and ravage Israel.69
66 See Het probleem, 65-66, for details.
67 See J. Ridderbos (De Psalmen 2.252-54, 303-4), who rejects a Maccabean date. J. A.
Goldstein's theory that Alcimus was the author of Psalm 79 places the psalm far too late
(I Maccabees [AB 41; Garden City, NY: Doubleday, 1976] 332ff.). Aside from the objections of
Ridderbos to a Maccabean date (2.252-54), it should be noted that if the psalm is this late,
it is impossible to explain the wrong authorship title known from LXX and MT Mss. Delitzsch
also places these psalms in Maccabean times (Psalms 2.325-28).
68 The concentration on the taking away of the golden shields in the historical accounts
seems designed to indicate the loss of Davidic glory (cf. 2 Sam 8:7).
69 Hence the reference to
Jacob in Ps 79:7. See E. H. Merrill,
History of Old Testament
Psalm 74 seems to have been written shortly after the looting of the
temple and possibly while Shoshenq
was still attacking
asks how long the Lord will continue to reject his people. We learn that in
Asaph, as we would expect, takes a cultic view of Shoshenq's destruction of
Lev 23:2) in the land.
In Psalm 79 Asaph again bemourns the defilement of the temple. It is
noteworthy that he here speaks of nations (pl.). That Shoshenq employed
soldiers from other nations is confirmed by the biblical account (cf.
2 Chr 12:3). Similar themes to Psalm 74 sound forth here. In v. 8 Asaph
asks that Yahweh not remember the iniquities of the chiefs (so Mynwxr, not
"forefathers" here) which brought about this invasion (cf. 2 Chr 12:1-2).71
There is also an echo of Ps 44:14 in v. 4. Psalm 44 would have been well
remembered by Asaph if, as argued, it dates to his lifetime (see above and
If this interpretation is correct, then the author of these two psalms may
well be the Asaph of Davidic times. Assuming he was around twenty years
of age at his appointment to temple service (cf. 1 Chr 16:4-5), and that this
took place in about 977 BC,72 then Asaph would have been about 71 at the
time of Shoshenq's invasion.
Psalm 83 is at once more and less of a problem. According to J. Ridderbos
it has been dated all the way back as far as Saul, and as late as Maccabean
times, with many suggestions in between.73 The difficulty in dating the
psalm results from the mention in it of places or peoples both ancient and
more recent (e.g., Amalek
Asaph does not seem impossible.
Given, then, that the eleven Asaph psalms of book 3 date no later than
Rehoboam's reign, it would seem possible that this collection was produced
by the men of Hezekiah. Hezekiah's reign was one of great religious reform.
Within that reform was an evident concern to gather and preserve impor-
tant material (revelation) from the Davidic and Solomonic period (cf.
70 Note that in v. 3 tvxwml should be read in conjunction with tvxvwml of Ps 73:18 as
"deception" and not "ruin"; cf.
Testament (Grand Rapids: Eerdmans, 1971) 218, and LXX, which translates e]pi> ta>j u[perh-
fani<aj. Perhaps Shoshenq deceived the priests, gaining entry for spoil, and then proceeded
to wreck the sanctuary.
71 In v. 1 the word yf occurs. This word only appears four times in the OT and is usually
taken to mean "ruin" (Jer 26:18; Mic 1:6; 3:12; Ps 79:1). However, it is notable that the LXX
in every case translates with on o]pwrofula<kion, that is, the "hut of a garden watcher" (see LSJ,
s.v.). Thus at least three (presumably) different translators at (presumably) different times
interpreted yf as a synonym of hnvlm. Such consistency is striking and suggests that this could
possibly be a meaning closer to ancient Hebrew usage. Psalm 79 therefore may not imply a
destruction of the city of
72 See Merrill, Kingdom, 244.
73 De Psalmen 2.330f.
THE DIVISION AND ORDER OF THE PSALMS 237
Prov 25:1).74 Hezekiah's concern for purity of worship even extended to the
words to be used in the temple service, for in 2 Chr 29:30 he and the officials
(MryWh) ordered the Levites to praise Yahweh specifically with the words of
David and Asaph the Seer (hzHh Jsxv dyvd yrbdb). This seems to imply that
some sort of gathering work had gone on to provide the suitable texts.
Possibly the collections of books 1 and 2 also came about at this time.
Ps 137:3 could well imply the existence in pre-exilic times of such a col-
lection of songs. Here in exile, captors demand that some of the temple
musicians sing Nvyc rywm. ryw here appears to be collective with partitive Nm.
Could this have been a technical term for a canonical collection?
The rest of book 3 exhibits no evidence of any earlier collection. Psalm 89
clearly shows that a date sometime in the exile or thereafter is demanded
for the book as a whole.
VI. Books 4-5
1. Division of Individual Psalms
When we turn to books 4 and 5 the situation is at once more confused and
more clear. There is here a much lower number of superscripts in general.
At the same time several Hebrew manuscripts combine quite a number of
psalms, especially between 90 and 99, and 114 and 119. These combinations
are often awkward and in nearly every case almost certainly secondary.75
I shall discuss only two cases of alternate division where the evidence seems
In the first place there is much evidence that Psalms 114 and 115 were
originally combined. Many Masoretic (Hebrew) Mss combine them (including
codex Leningradensis),76 as well as 4QPs°, LXX, Theodotion, Syriac, and
Jerome. Although internal evidence is not decisive, the praise of God and
ridicule of idols in Psalm 115 does flow well from the confession of deliverance
In the second place, as noted above, Psalm 147 is divided into two psalms
in the LXX. This psalm does easily fall into two sections that may have
been two independent psalms (vv. 1-11 and 12-20); however, there are no
Hebrew mss extant supporting this division. Thus the separation in two
must remain no more than a possibility.
2. Arrangement of Books 4-5
Book 4 begins again, as I have noted, with a change of authorship to help
mark the boundary line. Psalm 90 is thus a psalm of Moses. There appears
74 Of interest is t. B. Bat. 15a: "Hezekiah and his company wrote [btk] Isaiah, Proverbs,
the Song of Songs and Ecclesiastes." The verb btk here obviously signifies some kind of
copying or editorial work. The passage is cited in Young, Introduction, 202.
75 For a discussion of
some of these cases see
to be quite some significance to placing the psalm of Moses at the beginning
of this book, as his name appears no less than six times in book 4 (Ps 90:1;
99:6; 103:7; 105:26; 106:16, 23, 32). Outside of these instances the name of
Moses appears only once in the psalter (Ps 77:21).
The ensuing nine psalms seem to form a deliberate block introduced by
Psalm 91. This block is interesting for several reasons. First, only Psalm 92
has a superscript in MT, but the LXX shows that there were also super-
scripts for each one of these psalms. Second, many of these psalms have been
reckoned among a modern genre called Yahweh malak psalms. Although the
classification of this genre has been shown to be complex, modern study has
drawn attention to the similarity in content within this set of psalms.77
Psalms 92-99 all contain the theme of God's kingship and enthronement.
Besides this they have many other thematic elements in common (e.g.,
holiness of God, singing Yahweh's praise, etc.). An examination of them
reveals that together they celebrate Yahweh's victory in conquest over enemies
and thus his continuing rule.78 Several of them are given historical super-
scripts by the LXX. These reinforce the idea that they are victory psalms.
Psalm 93, for instance, has the LXX superscript: ei]j th>n h[me<ran tou?
prosabba<tou, o!te kat&<kistai h[ gh?: ai#noj &]dh?j t&? Dauid.79 Delitzsch
interprets this in line with a Talmudic tradition indicating the sixth day of
creation.80 The Talmud refers the second phrase to the populating of the
earth on the fifth day. This seems unlikely in view of the similar superscript
to Psalm 97, which is a clear reference to monarchial land in the kingdom
victory in the conquering and settling of land under David. The theme of
victory in conquest is a common factor in all the so-called Yahweh malak
psalms. Yahweh is celebrated again as king after the victory. There are
several military associations in the psalm. A study of the expression "to gird
oneself with strength" shows clearly that this is battle terminology (cf.
1 Sam 2:4; 2 Sam 22:40; Ps 18:33, 40; 65:7; Isa 8:9). Delitzsch argues
further that the rivers that are lifted up represent the mighty rivers of the
surrounding nations.'' However, Yahweh is more mighty than they (v. 4).
Psalm 96 is also considered a Yahweh malak psalm (cf. v. 10). Here too a
military victory seems to be celebrated. The reference to singing a new song
usually indicates the desire to celebrate a new victory attained through the
77 For complexity, see J. D. W. Watts, "Yahweh Malak Psalms," TZ 21 (1965) 341-48.
subjective nature of attributing psalms to this
so-called genre. For similarity, see
78 It is interesting to note that a further example of this genre (?) appears in Isa 42:10-13.
Again here all the common themes are present.
79 "For the day before the sabbath, when the earth/land was settled; a song of praise of
80 Delitzsch, Psalms 3.74. The reference is to Ros Has. 31a. See also Briggs, Psalms 2.299.
81 Delitzsch, Psalms 3.75-76.
THE DIVISION AND ORDER OF THE PSALMS 239
help of Yahweh. Deliverance from enemies is indicated (v. 2) and again
victory over their gods is celebrated.82
Of interest is the fact that Psalm 97, being like Psalm 93 a so-called
Yahweh malak psalm, also has a LXX superscript relating to the land.
It reads: t&? Dauid, o!te h[ gh? au]tou? kaqi<statai.83 A similar occasion to
Psalm 93 seems to be the background here. Yahweh's enemies have been
routed and have seen his glorious victory (v. 3, 6). The defeated are taunted
for their worship of idols (false gods) in the wake of the victory of the now
exalted Yahweh (vv. 7-9). Vv. 10-12 form a parenetic conclusion that bases
itself on the demonstrated faithfulness of Yahweh in preserving and
delivering his people.
Many of the psalms from this group (i.e. 91-99) are ascribed to David by
the LXX (Psalms 91, 93, 94, 95, 96, 97, 98, 99).84 These psalms themselves
bear no indication that Davidic authorship is unlikely. In fact, Psalm 99
seems to offer positive internal evidence of this period by its constant allu-
sion to the ark of the covenant.85 The ark is seldom heard of after the
division of the kingdom (but cf. Jer 3:16; 2 Chr 35:3). The fact that this is
a victory psalm after battle argues for David's and not Solomon's reign (cf.
1 Chr 28:3). That the ark was used by David to symbolize Yahweh's pres-
ence in battle seems clear from 2 Sam 11:11 (cf. Num 10:35-36).86 In
conclusion, it seems likely (given both internal similarities and LXX
authorship ascriptions) that all the victory psalms given here date from
the time of David.
It is tempting to see this series of victory psalms as an answer to the
questioning contained in the last psalm of book 3. There in the wake of the
destruction of the kingdom and capitulation of the Davidic king, the psalmist
cries to Yahweh to remember his covenant with David. Whilst this con-
nection is possible, and has indeed been argued,87 it seems to me rather
doubtful. As already argued, book 4 is considerably later than book 3, and,
in addition, Psalm 90 does not immediately answer the question anyway,
but contains a plea for the Lord's return itself (though not based on the
82 For an interpretation of the LXX superscript see n. 18 above.
83 "Of David., when his land is established."
84 With regard to Psalm 95, cf. Heb 4:7. With regard to Psalm 96, the parallel in 1 Chr
16:23ff. indicates that this psalm probably also dates from this period.
85 E.g., v. Mybvrk bwy (cf. 1 Sam 4:4; 1 Chr 13:6); v. 4 jlm zf (cf. Ps 132:8; 78:61; 2
Chr 6:41); v. 5 vylgr Mdh (cf. 1 Chr 28:2); v.7 vytdf (cf. Exod 25:22); note also the fact that
the ark contained the tdfh tHl (cf. Exod 31:18; Deut 10:5).
86 M. H. Woudstra has attempted to circumvent the prima facie interpretation of 2 Sam
11:11 (The Ark of the Covenant from Conquest to Kingship [International Library of Philosohy
and Theology: Biblical and Theological Studies;
119-21). However, I am not yet convinced that the passage does not imply that the ark was
out with the military. See C. J. Goslinga, Het Tweede Boek Samuel (Commentaar op het Oude
Testament; Kampen: Kok, 1962) 206.
The last section of book 4 contains a number of psalms of praise often
called Halleluyah psalms after their characteristic first and last lines. Strictly
speaking, Psalms 105 and 106 fall into this category. However, Psalms 103
and 104 also seem to be related. In the first place, the latter two are
connected by their common beginnings and endings (hvhy-tx ywpn vkrb).
Psalm 104 also connects to Psalms 105 and 106 by its concluding hy-vllh.
It may be that Psalms 103 through 106 are intended to be the answer to
Ps 102:19ff., especially 19b: hy-llhy xrbn Mfv.
The boundary between book 4 and book 5 is much more fluid than any
other. There is no authorship change here; in fact, both the last psalm of
book 4 and the first psalm of book 5 lack any superscript. Furthermore,
Psalm 107 seems to be closely related to the preceding by its opening verse.
of a new section after a group of Halleluyah psalms.88 Whilst it is noteworthy
that such a Todah psalm does frequently close a group of Halleluyah psalms,
the suggestion that it begins a new section seems to me a little far fetched.
The only real ground for it is the placement of Psalm 107 at the beginning
of book 5. Is not Psalm 105 a Todah psalm? Do Psalms 118 and 136 really
begin a new (sub)section?
A related grouping to Psalms 103-107 is 111-118, however, this does not
seem to follow on from any specific suggestion in the previous psalm. The
final Halleluyah group (Psalms 146-150) does, however, seem to follow from
Ps 145:21, the last of a series of Davidic psalms (138-145).89 The only
definite grouping of psalms that may have had a previous history as a
collection would appear to be the psalms of ascents, usually taken as a
group of pilgrimage psalms.90 For the rest, apart from sporadic indications
of relation here and there, book 5 does not bear any signs of deliberate
The fact of the fluidity between books 4 and 5, and the common use of
a group of Halleluyah psalms to answer to a previous psalm, would seem to
indicate that the two books were compiled together. Psalm 137 provides a
terminus a quo in the exile; however, the distinction between books 1-3 and
4-5 in the mss tradition noted above in the section on "evidence" would
seem to indicate a later date.91 Possibly these books were compiled during
88 Ibid., 187-90.
89 See Wilson, who aptly remarks: "In this context, Psalm 146 . . . represents the response
of David himself to the first half of 145:21. In
12, 19-20). The infectious praise spreads ever further in Psalm 148 where the angelic hosts and
the creation break forth into song (vv. 2-3, 11-12). In 149, focus returns to the people of
150, we hear the great hymnic answer to the second half of 145:21, toward which the whole
hallel has been building" (Editing, 194). Cf. Niemeyer, Het probleem, 154.
90 Niemeyer has shown that these also exhibit an inner connection (Het probleem, 59ff.).
91 Of interest is that
of Post-exilic Psalms by Means of Linguistic Criteria [in Hebrew;
Magnes, 1966]), whose
THE DIVISION AND ORDER OF THE PSALMS 241
the time of the literary work done by Nehemia, who also appears to have
amassed a considerable library (2 Macc 2:13).92 The number of Davidic
psalms in books 4 and 5 shows that there was still a considerable number
of psalms preserved from pre-exilic times.93
In summation we have seen that the division of the psalter into five books
is indeed not only warranted, but gives evidence of a historical develop-
ment of compilation over the ages since the times of Hezekiah or earlier.
This work of compilation into known and well used canonical collections
was probably completed only after the exile, perhaps in the time of Nehemia.
Whilst there are indications of internal ordering here and there, there
appears to be no systematic attempt to structure the psalter internally.
Given the historical development of compilation, the old interpretation of
midrash tehillim (on Ps 1:5) that the five books reflect the five books of Moses
is probably no more than a late reflection.94 The Sitz im Leben of this long
process of compilation appears to have been the need to furnish recognized
collections for use in the temple liturgy (cf. the connection to the liturgical
reforms of Hezekiah, 2 Chr 29:30, and the implications of Ps 137:3).95 The
psalter thus remains for us today as the church's prophetic songbook, teaching
her in inspired words of the greatness of her God, his wonderful deeds of
salvation, and his faithfulness to his covenant promises. It enables the
church to give back to God her prophetic response of praise that is his due.
2231 AR, Rijnsburg
studies have concluded that "the ten post-exilic Masoretic pss [by his standards] are all in the
last third of the psalter."
92 The song of praise in 1 Chr 16 has often been used as proof that books 1-4 of the
psalter were already in existence when Chronicles was written (see, for example, P. Skehan,
Ephemeridum Theologicarum Lovaniensium 46;
167-68, cited in
over portions of Psalms 96, 105, and 106. In light of my comments above on duplication in the
psalms, such an argument does not hold any weight. The composition in 1 Chronicles 16 could
easily be very ancient, even Davidic (recall that all examples of duplication in the psalter occur in
93 Compare 11QPsa "David's Compositions," which lists a total of 4,050 songs of David
itemized according to genre. These may no longer
have existed in the time of
list does show that there was probably once a considerable library of them.
94 The midrash itself is dated from the ninth century or later; cf. Hippolytus (d. 235), On
the Psalms, preface, for a similar view.
95 Thus I reject
counter both to its historical development and
to its purpose.
his notion of the purpose of the psalter by saying, "this is a collection to be read rather than
performed." If anything, I would conclude the opposite, though of course the reading of the
psalms is also highly profitable!
This material is cited with gracious permission from:
report any errors to Ted Hildebrandt at: