Guidelines for the Morphological Analysis of the LXX


The following guidelines were used by the staff of Computer-Assisted Tools for Septuagint Studies (CATSS) in preparing the Morphologically Analyzed Septuagint.  

A) USING HATCH AND REDPATH: 

      In general, Hatch-Redpath is considered the final arbiter for 
questions of morphological analysis (there are exceptions, however, 
outlined in what follows).  Thus many problems which are likely to arise 
can be settled by consulting Hatch-Redpath. These include questions of 
a) spelling; b) determining the lexical entry of a word; and c) parsing. 

     a) Spelling: 

     Frequently, the morph program errs in its spelling, since the
stems were based on classical literature. We are correcting these 
entries as they arise, but many others have yet to be detected. When
you find inconsistencies in orthography, make sure they are corrected in
line with Hatch and Redpath, not Liddell and Scott. 
     If Hatch and Redpath give several different spellings, 
choose the first. This is an arbitrary but necessary expedient. 

    b) Using H-R to determine the lexical entry:

    There are numerous cases in which it is difficult to decide what
the lexical entry should be. Here are a few examples which occur
repeatedly: 1) TI\ (meaning why) Is this from TI/ or TI/S? 2) OU(= (meaning 
where). Should this be derived from OU(= or O(/S? 3) OU(/TWS (meaning 
thus). From OU(/TWS or OU(=TOS? 4) I)DOU/ (meaning behold). From I)DOU/ 
or O(RA/W? 5) KU/KLW| (meaning around). From the preposition KU/KLW| or from KU/KLOS? 
    To settle these and related problems you should consult H-R to
see how these words are handled. Note, for example that I)DOU/ is 
a separate entry from O(RA/W; KU/KLW| is under the entry KU/KLOS; 
OU(= is under the entry OU(=; OU(/TWS under OU(/TWS; etc. 

   c) Using H-R to solve parsing problems:

    Morphological analysis can often be extremely difficult because of the 
inherent ambiguities of some words.  Take, for example, the words already 
cited above. How should they be handled? Should KU/KLW| be parsed as a 
preposition when it is used as a preposition, or as a DSM noun from 
KU/KLOS? No clear-cut answer can be given to cover all cases. The most one 
can strive for is consistency.
    Again, H-R can be of some limited assistance in the analysis. Since H-R 
(as against Arndt and Gingrich, which has KU/KLW| as a separate lexical 
entry) lists KU/KLW| under KU/KLOS (see above, b), this would mean that 
it is a declinable form from that noun. Thus, since we have opted to adhere 
to H-R, your analysis should look like this: 
KU/KLW|  N2  DSM  KU/KLOS, (not as a preposition).
    On the other hand, take the form I)DOU/. Note that since H-R
lists this as a separate form from O(RA/W, I)DOU/ should not be 
treated as an imperative from O(RA/W, but as an interjection. Your
analysis should, therefore, look like this: I)DOU/  I   I)DOU/ (the 
'I' should go in the type field). 


B) GENERAL CONSIDERATIONS 

    For those cases in which H-R does not prove helpful, the
general rule of thumb is to limit yourself as narrowly as possible to 
"morphological analysis" and not worry excessively about the syntactic
problems of the word. Of course, it should be recognized that there are 
instances in which it is necessary to make some decision about a word's 
function in a sentence before you can make a proper identification. 
Some of the cases in which this is necessary include KAI/ (adverb or
conjunction); E)NANTI/ON (preposition or adverb); E)NW/PION (preposition
or adverb). You should examine the sentence and make your decision
accordingly.

Order of non-alphabetic symbols. These symbols should go in the 
following order: 1) breathing; 2) accent; 3) iota subscript; and
4) diaeresis. 


C) THE "FIELDS" OF THE MORPHOLOGICAL ANALYSIS FILE

     1) Lemma (text) field (left column): 

     This column should list the word as it appears in
Rahlfs' text. All punctuation and all extraneous signs (brackets, 
parentheses, etc.) should be removed. (Punctuation should be removed
only at the very last stage of correction.) Words which are capitalized 
in Rahlfs' text (due to beginning a quotation, etc.)
but which are not proper nouns (according to H-R) should
have the asterisk removed, and their breathing and accent marks 
should be rearranged as for a non-capitalized word. 

     2) Dictionary field (right column):

     In general, each word should appear in the dictionary column in the
form in which it is listed in Hatch-Redpath, except that verbs
are listed in PAI1S rather than infinitive form, and verbal prefixes
are separated from the root.  (Several other exceptions are noted in what 
follows.) An especially significant variation from Hatch-Redpath
occurs in the treatment of proper nouns: see section D 1).
N.B.: Verbs are the only words which should have "prefixes" 
separated in the main dictionary entry. 

     3) Type and parse codes: 

     The "typing" and "parsing" of a word should conform to the 
form of the word listed in the dictionary field (see above).
The general principle is that a word is "typed" as it is used in
the text, rather than by some preconceived notion of its "ideal"
type or by an original or prototypical "type" (e.g. as it is used 
in Attic Greek style). Type and parse coding is outlined on an
accompanying sheet. Consult the creed for possible complications
such as treatment of numbers, substantives, and adverbs.
For the parsing of proper nouns, consult section D 1).


D) SPECIFIC CATEGORIES OF WORDS 

     1) Proper Nouns

     Here a substantial deviation from Hatch-Redpath occurs.
     The lemma (text) forms of proper nouns should be left as in Rahlfs'
text. The form entered in the dictionary column should Dconform to that
suggested by Rahlfs' text. THAT IS: If the textual lemma form 
in Rahlfs' text (and thus in the lemma column of the morph. analysis) 
1) is accented, 2) has a breathing mark, and/or 3) is inflected,
provide for these features in the dictionary column. If the 
textual lemma form lacks any of these features, omit them from the
dictionary column. Breathing marks and accents (in that order)
which go with the initial letter(s) of a proper noun should 
be placed after the '*' and before the first letter of the noun,
as in ordinary Greek usage. 
     BPARSING OF PROPER NOUNS: If the textual lemma form is inflected
(and the dictionary column thus records a Nom. Sing. form), 
parse the word as you would any other noun. If it is not inflected, 
place 'N' alone in the type code, and parse as fully as possible. 

     2) Adverbs 

     Adverbs are normally listed in H-R separately, under their own 
form. Since the MORPH program often derives adverbs from an 
adjectival or other form, output must be corrected in 
accordance with H-R. Provided that H-R lists the adverb separately, it
should be "typed" only as 'D', and the adverb itself should be
listed in the dictionary column. "Prefixes" should never be 
separated from an adverb in the dictionary column.

     3) Prepositions

     Any word clearly used as a preposition in syntactical
function (regardless of its lexical or grammatical derivation), 
and listed in its prepositional form in H-R, should be "typed"
only as 'P'.

     4) Numbers:

     All indeclinable cardinal numbers should be parsed only as 
'M' (in the type field). Declinable numbers should be treated as
any other adjective. The only declinable numbers between 1 and 20 
are EI(=S (1), TREI=S (3), and TE/SSARES (4). EI(=S is A3 in
masc. and neut. forms and A1A in fem. forms. TREI=S and TE/SSARES 
are A3. 
     The lexical forms EI(=S, E(/N, and TREI=S are parsed by the
MORPH program as 'M' and should be corrected to A3. 
     The "indeclinable" number 2 (DU/O) does have several declinable forms
[mainly DUSI/(N)]. DU/O should be typed only as 'M'. DUSI/(N) should be 
typed as 'M,' and should be fully parsed as well (as plural, not dual). 

     5) Substantives: 

     On the question of substantives and whether to parse them as nouns 
or adjectives: If an adjective is consistently used as a
noun, parse it as a noun. An example of this is E)/RHMOS, which 
LSJ identifies as an adjective. However, since this is consistently 
used as a nominal form in the LXX, it should be parsed as a noun. 
The same applies to E(RPETO/N (N2N, not A1) and PETEINO/N 
(N2N, not A1).

     6) Adjectives and adverbs of degree (compar. and superl.): 

     a) In fourth column of parse field, indicate 'C' or 'S' if 
it is an irregular form.
     b) Check Hatch-Redpath to see if comparative or superlative form 
is a separate entry. If so, list the comparative or 
superlative form itself in the dictionary column. 


    To assist in the morph process, we include here a description of
problem words, and the ways we have solved them.


(*THIS FILE CONTAINS A LIST OF PROBLEM WORDS IN THE MORPH ANALYSIS; PLEASE
FEEL FREE TO CONSULT THIS FILE OR ADD TO IT WORDS THAT ARE CREATING 
HARDSHIP, as long as such additions are approved by the council of
elders.*) 

I.] GENERAL NOTES (Arranged alphabetically by key word):


ADJECTIVES and adverbs of degree (comparative and superlative): 
   see above, section D 6). 

When Hatch-Redpath gives ALTERNATE forms of a word, such as 
    BORE/AS, BORE/HS, BORRA=S, use the first choice as the dictionary 
    form, even if one of the others is closer to the form actually found
    in the text you are parsing. The following are exceptions:
       a) OU(/TWS  (dict. form should be OU(/TWS, not OU(/TW contra H-R). 
       b) PROPER NAMES containing an I/EI variant (we, with Rahlfs, 
          use the I form).
       c) E)SQI/W, not E)/SQW 
       d) A)RGUROU=S, not A)RGU/REOS

ASTERISKED words: In the case of proper nouns, make sure
    not to remove the asterisks from either column. All other words 
    (verbs, common nouns, etc.) should have asterisks removed from both 
    columns [see sections C 1) and D 1)]. 

CRASIS: 
                KA)GW\          C+RP NS        KAI/ + E)GW/ 
                KA)\N           C+X            KAI/ + A)/N

MIDDLE vs. Passive: 
    Always distinguish these, even though the MORPH program parses
    all ambiguous forms as Middle.

Prepositional PREFIXES for verbs--List the prefix separately only 
    if the unprefixed verb is attested as a form in its own right 
    (consult unabridged Liddell-Scott). 

The following words, while not adhering strictly to H-R, should not be
    changed, but should remain as the PROGRAM analyzes them:
             --MI verbs 

All PROPER NOUNS should be fully parsed, as far as possible, in 
    case, number and gender. If any of these are unclear, leave a blank.
    (See section D 1).) 

Examples of SUBSTANTIVES: 
         E(RPETO/N    N2N   parse   E(RPETO/N 
         PETEINO/N    N2N   parse   PETEINO/N 
         E)XQRO/S     N2    parse   E)XQRO/S
         E)/RHMOS     N2    parse   E)/RHMOS

    HOWEVER, the following words should be parsed as ADJECTIVES:
         A)/RSHN
         QH=LUS 
         H(/MISUS 

SUBSTANTIVES: See also section D 5).

For treatment of the various "types" of THIRD DECLENSION nouns and
    adjectives (N3E, N3I, A3C, etc.), see the appended materials. 


II.] SPECIFIC WORDS (Arranged in Roman alphabetical order): 


A)F' OU(= .  See under OU(= . 

Forms of A)MFO, A)MFO/TEROS, &c., are parsed by the program as A)MFW. 
    This should be corrected to:
               A1A     appropriate parse   A)MFO/TEROI    (see H-R).

A)NA\ ME/SON: ME/SON should be neuter (see BAG, s.v. ME/SOS, #2): 
                   A)NA\     P          A)NA/ 
                   ME/SON    A1  ASN    ME/SOS
    Similarly:  E)K ME/SOU    and    E)N ME/SW| 

Prefer:   A)RGUROU=S  to  A)RGU/REOS
          XALKOU=S    to  XA/LKOUS
          XRUSOU=S    to  XRU/SEOS. 

A)RGUROU=S .  See also under ALTERNATE. 

A)/RSHN .  See under SUBSTANTIVES VS. ADJECTIVES. 

TO\ DEU/TERON: Frequently, this expression is adverbial, meaning
   the "second time." Both forms should, however, be parsed as
   ASN. 
       TO\       RP     ASN    O( 
       DEU/TERON A3     ASN    DEU/TEROS

DU/0 .  See above, section D 4).

E)/GGISTA-- parse as adverb from E)GGU/S: 
              E)/GGISTA          D            E)GGU/S 

E)NANTI/ON and E)NW/PION: 
    In order to analyze these forms, you must determine whether they
    function as prepositions or adverbs. Since the program always parses
    them as adverbs, you may have to make the necessary corrections.
    When these forms are prepositions, the analysis should be:
             E)NANTI/ON        P                E)NANTI/ON
             E)NW/PION         P                E)NW/PION 
    The same applies to E(/WS, O)PI/SW, E)/NANTI, PLH/N, PE/RAN, E)/CW, 
    O)/PISQEN, U(PERA/NW, U(POKA/TW and U(POKA/TWQEN. When used as
    prepositions indicate only with 'P' in type column. 

E)PA/NW: This word is sometimes used as a preposition, sometimes
             as an adverb: parse accordingly. 

E)/RHMOS .  See under SUBSTANTIVES VS. ADJECTIVES.

E(RPETO/N .  See under SUBSTANTIVES VS. ADJECTIVES. 

E)RW= : All forms should be parsed under EI)=PON in dictionary column.
        Similarly for its derivatives, EI)/RHKA and E)RRE/QHN.

E)SQI/W .  See under ALTERNATE. 

E(/WS OU(= :
          E(/WS         P          E(/WS
          OU(=          RR  GSM    O(/S  (agreeing with implied XRONOU).
    The same applies to  OU(=  in  A)F' OU(= .

E)XO/MENA: When this word is used in "frozen" form as a preposition,
             parse as    V1   PMPAPN   from    E)/XW

E)XQRO/S .  See under SUBSTANTIVES VS. ADJECTIVES.

H(/MISUS .  See under SUBSTANTIVES VS. ADJECTIVES.

I)DOU/ :
    When it means "behold," it should be parsed as an interjection (I), 
    following H-R. The output should be corrected, since the program
    consistently parses this word as an aorist imperative from O(RA/W.
                  I)DOU\          I             I)DOU/

KA)GW\ .  See under CRASIS. 

KA)\N .  See under CRASIS.

KAI/ :  Be sure to distinguish between conjunctival and 
   adverbial use. If it is an adverb, simply put D in the type field
   (not C in the type field and D in the parse field).
   a) Also note that with a subordinate construction (especially
      when the subordinate clause is introduced by E)A/N),
      KAI/ will frequently introduce the main clause. 

KAQA/: Program declines as a preposition [KATA/ + O(/S], but this 
   word is more correctly parsed as an adverb. The same applies to
   KAQA/PER, KAQO/TI and KAQO/. Thus: 
            KAQA/          D         KAQA/
            KAQA/PER       D         KAQA/PER 
            KAQO/TI        D         KAQO/TI
            KAQO/          D         KAQO/

KU/KLW|: Although this is sometimes used as a preposition or adverb,
            simply give it as N2  DSM from KU/KLOS (pace H-R).

MAKRA/N: When used adverbially it should be marked: 
                  MAKRA\N         D         MAKRA/N      (see H-R). 

A)NA\ ME/SON: ME/SON should be neuter (see BAG, s.v. ME/SOS, #2): 
                   A)NA\     P          A)NA/ 
                   ME/SON    A1  ASN    ME/SOS
    Similarly:  E)K ME/SOU    and    E)N ME/SW| 

Watch for O(/, which is frequently misparsed by the morph program 
    as a masculine relative pronoun (should be neuter). 

O)/LLUMI or O)LLU/W--choose the latter. 

O(RA/W and OI)=DA: A slight deviation from H-R occurs here. 
    All words under EI)DEI=N in H-R should be parsed under the dictionary 
    form OI)=DA (these will be Perfects or Pluperfects). For
    forms which H-R has under O(RA/W, continue to use the same
    word, i.e. O(RA/W. For forms in H-R under I)DEI=N, parse under
    O(RA/W (these will be 2 Aorists). 

OU(=, from O(/S (meaning where): Since H-R has this under 
    a separate lemma [OU(=], it should be parsed as follows:
             OU(=              D                OU(=
   [N.B. In H-R, this applies only to OU(= when it means "where". 
   When OU(= has the adverbial meaning of "when" --as for example,
   with E(/WS--then it should be parsed as RR GSM from O(/S.] 

E(/WS OU(= :
          E(/WS         P          E(/WS
          OU(=          RR  GSM    O(/S  (agreeing with implied XRONOU).
    The same applies to  OU(=  in  A)F' OU(= .

OU(/TWS: Although the program handles OU(/TWS as an adverb from 
             OU(=TOS, it should be an adverb from OU(/TWS (pace H-R): 
                   OU(/TWS        D             OU(/TWS 

OU(/TWS .  See also under ALTERNATE.

PANTOS in DIA/ PA/NTOS is masculine (sc. XRO/NOU).

PETEINO/N .  See under SUBSTANTIVES VS. ADJECTIVES. 

PLHSI/ON: Parse as an adverb (D), modified (usually) by a masculine 
          article.

PRO/TERON: Note that when it is adverbial, it is listed in H-R
    as a separate lexical entry. The following forms should be
    categorized carefully according to H-R: 
       PRO/TERON, PRO/TEROS, PRW=TON, PRW=TOS.

TO\ PRWI/+ :
                  TO\             RA   ASN      O(
                  PRWI/+          D             PRWI/+

QH=LUS .  See under SUBSTANTIVES VS. ADJECTIVES.

TI/ , in its meaning "why," should be listed as:
               TI\        RI   ASN           TI/S 
    because it is so listed in Hatch-Redpath. 

In I(/NA TI (sc. GE/NHTAI), TI/ O(/TI (sc. GE/GONEN), and 
    TI/ GA/R:  TI should be nominative. 

TI/S and TIS: These forms should be carefully distinguished (see H-R).
    The interrogative form (TI/S) always has an acute accent on the first 
    syllable; the indefinite (TIS) is either unaccented, or accented on 
    the second syllable (and is enclitic).

XALKOU=S .  See under A)RGUROU=S. 

XRUSOU=S .  See under A)RGUROU=S.