Routledge Ret-'ilJ(t/s

Morphophonemics of Modern Hebrew

Rti'intlr present a reissue of Noam Chomksy's 11A thesis, written in

Ro!ltlt,~r;e
1951, and first published in book form in 1979. Mmpbopbonemin ol Modan
Hdm·u• is a Lmdmark study in linguistics and generative phonology, which
provides not only an analysis of morphophonemics but of the entire grammar
of Modern Hebrew from syntax to phonology.

This work is of singular importance as it contains the genesis of the author's

work in the field of generative grammar which has had such a profound
impact upon the study of linguistics. This reissue of a truly pioneering work
will be of great interest to all those concerned with generative grammar <md
its origins, and with the progression of thought of one of the greatest minds
of our time.
Morphophonemicsof
ModernHebrew

NoamChomsky
Firstpublishedin1979
byGarlandPublishing,Inc
Thisedition firstpublishedin2011byRoutledge
2ParkSquare,MiltonPark,Abingdon,Oxon,OX144RN
SimultaneouslypublishedintheUSAandCanada
byRoutledge
711ThirdAvenue,NewYork,NY10017
RoutledgeisanimprintoftheTaylor&FrancisGroup,aninformabusiness
©1979AvramNoamChomsky
Allrightsreserved.Nopartofthisbookmaybereprintedorreproducedorutilisedin
anyformorbyanyelectronic , mechanical,orothermeans,nowknownorhereafter
invented,includingphotocopyingandrecording,orinanyinformationstorageor
retrievalsystem,withoutpermissioninwritingfromthepublishers.

Publisher’sNote
Thepublisherhasgonetogreatlengthstoensurethequalityofthisreprintbutpoints
outthatsomeimperfectionsintheoriginalcopiesmaybeapparent.

Disclaimer
Thepublisherhasmadeeveryefforttotracecopyrightholdersandwelcomes
correspondencefromthosetheyhavebeenunabletocontact.

ALibraryofCongressrecordexistsunderISBN:0824096886

ISBN13:978-0-415-68792-8(hbk)
ISBN13:978-0-203-35745-3(ebk)
 Morphophonemics
of Modern Hebrew

Noam Chomsky

Garland Publishing, Inc. • New York & London

1979
library of Congress Cataloging in Publication Data
Chomsky, Noam.
Morphophonemics of modern Hebrew.
(Outstanding dissertations in linguistics; 12)
A revision of the author's thesis (M.A.),
University of Pennsylvania, 1951.
Bibliography: p.
1. Hebrew language-Morphophonemics.
I. Title. II. Series.
Pj4600.C58 1979 492.4'5 78-66579
ISBN 0-8240-9688-6

© 1979 Avram Noam Chomsky

All rights reserved

All volumes in this series are printed on acid-free,

250-year-life paper.
Printed in the United States of America
 December, 1951 Revision of

Morphophonemics of Modern Hebrew

Avram Noam Chomsky

"A Thesis

In Linguistics

Presented to the Faculty of the

Graduate School of the University of

Pennsylvania in partial fulfillment of

the requirements for the degree of

Master of Arts"

June, 1951
TABLE OF CONTENTS

0. Introduction.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . l

1. Notation........................................ 6
2. Syntax. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

3. Further Notational Statement . . . . . . . . . . . . . . . . . . . . 24

4. Morphology and Morphophonemics . . . . . . . . . . . . . . . . . . 26

5. Justification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49

6. Sample Derivations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58

7. Omissions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65

FOOTNOTES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67

BIBLIOGRAPHY. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 4
o. Introduction

A grammar of a language must meet two distinct kinds of

criteria of adequacy. On the one hand it must correctly

describe the 'structure' of the language (i.e., it must

isolate the linguistic units, and, in particular, must dis-

tinguish and characterize just those utterances which are

considered 'grammatical' or 'possible' by the informant,

including as a special subclass those of the analyzed corpus.

On the other hand it must meet requirements of adequacy

imposed by its special purposes (e.g., pedagogical, as a

basis for comparative study, etc.), or, in the case of a

linguistic grammar having no such special purposes, re-

1
quirements of simplicity, economy, compactness, etc. Thus

the linguistic analysis of a language L can be described as

the process of determining the set of 'grammatical' or 'sig-

nificant' sentences of L (i.e., of determining the ex-

tension of the predicate 'grammatical in~·), or, in other

words, it is the process of converting an open set of sen-

tences--the linguist's incomplete and in general expandable

2
corpus--into a closed set--the set of grammatical sentences--

and of characterizing this latter set in some interesting

way. Accordingly we might distinguish and consider separately

two aspects of the linguistic analysis of a language, a

process of 'discovery' consisting of the application of

the mixture of formal and experimental procedures constituting

linguistic method, and a process of 'description' consisting

of the construction of a grammar describing the sentences

-1-
which we know from step one to be grammatical, and framed

in accordance with the criteria related to its special

purposes.

Although the distinction between the processes of

discovery and description is clear enough in the case of

grammars with special purposes, it is perhaps less clear

in the case of a linguistic grammar constructed solely

in accordance with considerations of elegance, since the

process of discovery itself can perhaps best be understood

as the process of constructing a tentative grammar speci-

fying the grammatical sentences by listing the linguistic

elements and their permitted arrangements on various levels.3

Furthermore it is clear that considerations of elegance are

operative in the original process of discovery, i.e.,

that they have a distinct place in the framing of the

procedures of linguistics themselves. Thus in setting up

such linguistic elements as morphemes (a process of dis-

covery) we must consider properties of the linguistic

elements themselves (e.g., perhaps minimization of their

number) and properties of the statements describing these

elements and their relationships (e.g., perhaps minimi-

zation of their number),

4 and the same is true on other

levels of linguistic analysis. This consideration amounts

to the requirement that the predicate 'grammatical in L'

(and in general, the procedures of linguistic analysis)

be defined and analyzed in a metalanguage to the language

in which grammars are written, and consequently in a

-2-
meta-metalanguage to the language ~under analysis. Thus
one of the considerations involved in setting up linguistic
elements in a particular way, and consequently, in determining
what are in fact the grammatical sentences, will be the total
simplicity of the grammar in which these elements appear.
However it will still be useful to con@ider the
processes of discovery and description separately. For the
most reasonable way to approach the investigation and analy-
sis of the notions of simplicity in terms of which 'gram-
matical in L' is defined (i.e., those notions of elegance
that are relevant to the very formulation of the proce-
dures of linguistics) seems to be to assume, for some lan-
guage, that the grammatical sentences are fixed (i.e.,
that the process of discovery has been completed) and to
determine the effect on grammar-formulation of explicit
considerations of simplicity imposed on the grammatical
statement.5

The outline of Modern Hebrew grammar given below is

an example of the second step in linguistic analysis,
artificially isolated. It is assumed that the sole pur-
pose of the grammar is to generate a closed body of sen-
tences, these having already been determined. Hence the
grammar must be designed in such a way as to be the most
efficient, economical, and elegant device generating just
these sentences.

The grammar consists of the following parts:

1. A syntactic statement giving permitted
arrangements of morphemes in sentences.

-3-
 2. A morphemic constituency statement giving
permitted arrangements of morphophonemes in
morphemes.
3. A series of morphological and morphophonemic
statements transforming any grammatical
sequence of morphemes into a sequence of
phonemes.
4. A phonemic statement (transforming phoneme
sequences into phone sequences). 6

The effect of the first two parts is to give the per-

mitted sequences of morphemes by presenting sequences of
'morpheme names', some of them in morphophonemic spelling.
The first part will only be sketched here, and the second 7
and the fourth will be entirely omitted. The third part
will be given in detail. Beginning with a sequence of
morphemes from parts one and two, each statement of the
third part of the grammar specifies certain changes which
must be undergone by any sequence of a certain shape.
It will appear that an order is imposed on the statements,
relative to certain criteria of simplicity. Thus the
statements are ordered so as to present a maximally simple
grammar. The actual demonstration of adequacy given
below must be taken in a limited sense only. What is shown
is that any single interchange of consecutive statements
will necessitate changes which increase the complexity of
the grammar. 8 Thus the simplicity of the system is at what
might be called a 'relative maximum' with this ordering of

-4-
statements. It is not excluded that some complicated set
of interchanges of the statements might give a simpler
grammar, or in fact, that a total recasting in different
terms might be more elegant. Thus this investigation is
limited in that only one 'dimension' of simplicity is con-
sidered, viz., ordering. Actually a complete demonstration
would have to show that the total simplicty is greatest
with just the given ordering, segmentation, classification,
etc.
For the formulation of any relatively precise notion
of simplicity, it is necessary that the general structure
of the grammar be more or less fixed, as well as the no-
tations by means of which is constructed. We want the notion
of simplicity to be broad enough to comprehend all those
aspects of simplicity of grammar which enter into conside-
ration when linguistic elements are set up. Thus we want
the reduction of the number of elements and statements,
any generalization, and, to generalize the notion of
generalization itself, any similarity in the form of
non-identical statements, to increase the total simpli-
city of the grammar. As a first approximation to the notion
of simplicity, we will here consider shortness of grammar
as a measure of simplicity, and will use such notations
as will permit similar statements to be coalesced. To
keep this notion of simplicity from reducing to an ab-
surdity, the notations must be fixed in advance, and must be
chosen to be neutral to any particular grammar, except
with respect to the considerations they are chosen to

-5-
reflect. 9
Given the fixed notation, the criteria of simplicity
governing the ordering of statements are as follows:
that the shorter grammar is the simpler, and that among
equally short grammars, the simplest is that in which the
average length of derivation of sentences is least.

1. Notation

The grammar, then, will be a set of transformation state-

ments each of which transforms a given representation of
a sentence into a more specific one. 10 If a, S, y, with
or without subscripts and primes, stand for any sequences
(or zero, henceforth 0) of the elements appearing in state-
ments (e.g., sequences of phonemes, morphemes, phrases,
etc., including brackets, dots, etc.), then the basic
transformation statements of the grammar will be of the
form:

(1) a~ S, where

where a and S contain no notational elements but are

simply sequences of the elements set up to represent parts
of sentences (phonemes, morphemes, etc.). This means that
a is transformed by this statement into S, when conditions
obtain.
If a=a s y and S=a 1 s1 ,y, we rewrite (1) as:
1 1

(2) sl~ sl' in environment al ___y, where ...

-6-
 The notational devices which will actually be used

should be introduced definitionally (by so-called 'con-

textual definitions') be describing a procedure to convert

each expression using these notations into a sequence

of simple expressions of the form (1) or (2) (which is

reducible to (1)) where no notational elements appear.

Two kinds of brackets--{},[]--and two kinds of parentheses--

(), <>--will be employed as follows:

Nl. an abbreviation for

(i) '··· a 1 (ii) ct2 •.. ' ' ... ' (n) ' •.. a n ••• ' '
-
I

in that order. If two sets of brackets with a different

number of rows appear, either can be expanded first. If

two or more sets of brackets of this form with the same

number of rows appear, then they are expanded simultaneously,

the kth row of the first concurring with the kth row of the

second. For example

stands for

(i) 'ala2a4~ sls2' where yl

1
(ii) ala3a4--'r sls3' where y2

in ~ order. To indicate how many rows a given set of

brackets have, '---' is written where no element a occurs.

-7-
Thus has three rows.

N2. Nl holds in exactly the same form for [].

N3. A statement containing one or more elements in main

parentheses () is an abbreviation for two statements, one in

which all of the parenthesized elements appear, and one in

which none of the parenthesized elements appear, in that

order. For example

stands for

with (i) preceding (ii), and (i) in turn standing for two

statements by the same process of development.

N4. A statement containing one or more elements in

parentheses <> is an abbreviation for the conjunction, in

any order, of all statements with zero or more of the

parenthesized elements omitted. For example

stands for

(ii) 'aSy--? a ' (iv)

l

-8-
taken in any order. Order does not happen to be important

in the statements of the grammar in which <> is used. But

i t could be, and an order could be imposed. Note that the

appearance of a single <> is just like that of a single

() (except that order is imposed in the second case).

It remains to give the interpretation for the cases

where several of these four notations co-occur in one

statement. To do this we have to give an order of priority,

stating which of co-occurring notations is to be expanded

first, so as to have a unique interpretation for each

statement. The order of development follows these two

principles:

N5. No brackets or parentheses are expanded if en-

closed within brackets or parentheses. I.e., at each step

in the development of a sentence only main brackets or

parentheses may be developed.

N6. If there is more than one set of main brackets

or parentheses, they are developed in the order (iJ {},

(ii) [], (iii) (), (iv) <>; i.e., in exactly the order in

which they were introduced by Nl-4.

This now gives us an explicit step by step procedure

for converting each statement of the grammar into an ordered

sequence of statements of form (l) or (2). Notice that the

case of co-occurring brackets of the same kind with the

same number of rows is analogous in interpretation to matrix

multiplication, while co-occurring brackets of different

kinds give essentially the Cartesian product.

-9-
 One other point concerning co-occurrence of various
notations needs clarification, namely, occurrence of
brackets and parentheses within other brackets or paren-
theses.

N7. Each set of brackets or parentheses is_ treated

as a single element when inside of a containing set.
E. g., the
main braoket in 'e[~ll r ha' two rows.

N8. In accordance with customary practice, a set of

brackets with a single row is used to give the membership
of a class. Thus {o: 1 , o: 2 , ... , o:n} is the class containing
as members o: 1 , o: 2 , ... , o:n. A statement of the form 'o:={o: 1 ,
o: 2 , ... , o:n}' is interpretable in terms of (1) and (2).
It can be taken as an abbreviation for 'o:~o:
1 or o:~o:
2
or or o:~o:n'· We write •o: 1 go:', •o: 2 go:', etc., to indi-
cate that o: is a member of o:, o: 2 is a member of o:, etc.
1
'o:i' will be taken to designat.e the ith member of o: con-
sidered to be ordered from left to right as given, i.e.,
o:i' and 'o:' will be used as a variable ranging over members
of the class o:. If o:i itself contains variables or brackets,
then any explicit expanded expression produced by developing
o:i is taken to be a member of o:. E.g., if

then o:lEo:, o:2o:3£o:, o:2o:4o:5go:, o:2o:4go:.

-10-
 As an example of the functioning of these rules of

development consider the following case:

(i)

The expansion is given step by step as follows, with

the rule governing each step.

(ii) a:
by Nl, priority given
to {} by N6.
b:

(iii) al: (al)a2~ (yl)y2

by N2, priority given
a2: a.3~ (yl)y3 } by N6.

bl: s~ (yl)y2
by N2, priority given
b2: by N6.
s~ (yl)y3 ]

(iv) ala:

alb:
}y N3.

a2a:
}by N3.
a2b:

bla:

blb:
}y N3.

}y
b2a:

b2b: N3.
B-4 y3

The final set is given as (iv), in that order. The use

-11-
of five pairs of brackets and parentheses eliminates four-

teen occurrences of a , S, ... ,etc., which may themselves

1
be long expressions.

2. Syntax

In accordance with the plan stated above, the syntac-

tic statement will be skeletal and incomplete. It is

intended merely as an indication of the framework into

which the detailed morphological statement fits, and as

a sketch of the general structure of the simple Hebrew

sentence. Syntactic and morphological considerations may

be interrelated in the process of grammar construction,

since we are seeking the simplest total set of transfor-

mations. Thus if the morphology were presented indepen-

dently of syntactic considerations, one might consider the

various forms of plural and feminine suffixes to be parts of

special vowel patterns, added to roots to make up stems.

But this would hide the characteristic feature of these as

long components, i.e., components of phrases or sentences

rather than of words. Such a formulation would greatly

complicate the syntactic statement. Consequently the syn-

tactic statement must be detailed enough so that all

further elaborations of it will be irrelevant to morphologi-

cal considerations.

Sl. Sentence~ Elementary sentence <Connective+Sentence>

Sl must be reapplied until 'Sentence' is eliminated.

-12-
 S2. Connective={v, aval, o, ... } [{and, but, or, ... }]

83. Elementary Sentence~

[:\:_=1, ... ,4]

The sentence is now represented in terms of phrases

(NP=Noun Phrase, VP=Verb Phrase, PP=Prepositional Phrase,

LCi is the :hth member of the class of Long Components, Lb

3
is the third person pronoun in its free form).

[j=l, ... ,4]

S5. PP--7 Preposition+NPLC~ [j=l, ... ,4]

Prepositions and their special forms before various

kinds of NP's are left out of consideration below.

S6. NPLCi=
-
{ [Lak]
-
LC.l '
Lb -
-k

where (i) [~,~=1, ... ,4], [k=l,2,3]

(ii)
i [::_~in env. LC~ .. · - ]

' does not extend outside of

the NP
12
(iii) o~ ha (or 0 in env. -1
N )

and all a's change simultaneously

and identically since o repre-
sents a long component.

-13-
 ha is the definite article, a long component in Hebrew.

Lak and Lbk are forms of the personal pronoun, Z is a

demonstrative, and S is a morpheme occurring automatically

with the first N of an N N 'compound noun'. This 'con-

1 1 1
struct state' (Hebrew--'smixut') construction seems to be

disappearing as a productive construction in colloquial

12
speech in favor of Nl sel N (sel=of). Since~ is
1
independent of 1, certain of the arrangements of LC's

are purely internal matters of the Noun Phrase. Notice

that when two LC patterns occur, the one falling on the

first element of the phrase carries the LC of the sentence,

just as with VP's (see S4).

38. [i=la, lb, 2]

+
M =MU{alb}

U+=UU{a 2 b}

-14-
 M and U are the major structural classes of words,

the units within which the morphological transformations

function. They correspond, as can be seen by their mor-

pheme class constituents, to Noun and Verb, respectively.

S9. + ~ # in env.

The units within # __# will be the words with which

the morphology is concerned. M, M+, U, and u+ will be used

throughout the morphological section to indicate within

which class of words a rule holds.

SlO.

R and VlP are the major morpheme classes, roots and

vowel patterns, respectively. Ft is the morpheme of the

future tense, Lc is the infinitive morpheme, u 0 and u2

are classes of verb affixes (FtEU , see below, Sl5).
2

Sll.l4

~-is the morpheme of present tense (~-EU 2 , see Sl5).

Thus present tense verbs can also be nouns. Ns is a class

of nouns, mostly foreign borrowings, best construed as

-15-
not composed of roots and vowel patterns. They are not

treated in the present study. N2 is a subclass of N ~

1
essentially, adjectives. A much more explicit charac-
terization of N could be given, but is omitted here.
2
M is a set of noun prefixes. The forms h- ... are formed
0
from the corresponding verb forms. h-EU 0 (see Sl5) and
the two vowel patterns given are the members of VlP
2
(see S20). F=Feminine.

Sl2. M is any sequence or non-zero subsequence of

1
morphemes of one of the following forms:

(i)

(ii) on+aY
3
(iii) on}+ {i ( +u(F+i) )]
{ an u(~+iT+u))

where M ~

and V--7 V:
1 !
u in env. [· ~- ·]
.

[V=Vowel]
l--:e
l ... in env. h-···
}
--

The general meanings of these suffixes are roughly

as follows: aY --dual; a:n and a:Y --agent, person con-

2 3
nected with; o:n--thing connected with; u:--abstract noun

formative; i:--adjectival formative. Thus most N's ending

in i: are in N 2 .

Sl3. LC={0, E, ~' PF},

where LCl.--7 {~'~or in env. a:Y2--

PF in env. u: __ , and
}
u:~ u:y in env. __ PF

and if one LCi--7 a, then all

-16-
 813 (cont. ) . instances of LC. in the sen-
tence---:; a.l5 :!:.
and i~ 0, sometimes, in env. __ P.

F is the Feminine, P the plural long component.

=
~3 in env. ha# , sometimes
814. z~
{ zot
{eyle
ze
II
11
II
11
F }
PF and !:., PF~

815. M ={0, ml-, m2,

0
!-}
u0 ={0, !:!_-, li_-, g-li_-}

U2={0, Ft, !!!_-}

ml-, m2-, and T- are noun prefixes, N- and h- are verb

prefixes. N-+Verb is generally passive, h-+one type of
verb stem (with VlP=a ... a, see 820) is generally causative,
g-+the second type of verb stem (with VlP=i ... :~) is
generally reflexive. Ft signifies future tense, ~- sig-
nifies present tense.

816. Lak~ 0 in env. m-+

La 2 --7 La I2 , sometimes, in env. #n ... Ft+ __ , where

# is the first # in the sentence,
n
and ... contains no#. La~ will be
the morpheme of the imperative, which
varies freely with second person
future in this position.

817. 0~ et in env. U# NP
[i.e., before the NP
which is part of the VP,
see S4],
when NP contains ha#, where et is a member of
the class of prepositions [see S5].
-17-
Except for R and VlP, sentences are now completely specified

in terms of morphemes. A few illustrations of the repre-

sentation of a sentence in terms of morphemes by development

of the syntactic statements may be helpful here. The number

of the Syntactic Transformation applied is in brackets

on the same line as the result of that transformation.

A. l. Sentence

2. Elementary Sentence [Sl]

3. NPLC2+VPLC2 [S3]
LC
4. NP 2+V +LC +V [S4]
1 2 2
5. La +LC +v +LC +V [S6]
2 2 1 2 2
6. v +La +LC +V
1 2 2 2 [37]
7. #V +La +LC ##V #
1 2 2 2
[S9]

8. #U +R+VlP +U +La +LC ##U +R+VlP +Ft+Lc#

0 2 1 2 2 0 2
[SlO]

9. " +F## " [313]

10. #R+VlP [315]

2 +La 2 +~##h-+~+VlP 2 +Ft+Lc

An example, filling in actual roots and vowel patterns,

might be "rad.t lhitraxec" ('you(F) wanted to get washed').

B. l. Sentence

2. Elementary Sentence [Sl]

3. NP 1 C4+VPLC4 [S3]

4. NPLC4+V +LC +NPLCl [S4]

1 4
5. N 1+LC +£+ha+N 1 +LC +ha+N +LC +v +LC +N
4 1 2 4 1 4 1
+LC +£+ha+N +LC [S6]
1 1 3
6. #N
1 +LC 4 +~#ha#N 1 +LC 1 #ha#N 2 Lc 4 ##V
1 #LC
4
##N +LC +~#ha#N [S9]
1 1 1 +LC 3 #

-18-
 7. #N
1 +LC 4 +~#ha#N 1 +LC 1 #ha#N 2 Lc 4 ##U 0 RV1P 2 U 1
La Lc ##N +LC +~#ha#N [SlO]
1 4 1 1 1 +LC 3 #
8. #M
0 RV1PM 1 Lc 4 ~#ha#N 8 Lc 1 #ha#M 0 RV1PLC 4 ##U
0
RVlP 2 U La Lc 4##M RVlPLC #ha#M RVlPLC # [Sll]
1 1 0 1 0 3
9. #M
0 RV1P+i+LC 4 ~#ha#N 8 LC 1 #ha#M
0 RV1PLC 4 ##U
0
RVlP U La Lc ##M RVlPLC #ha#M RVlPLC # [Sl2]
2 1 1 4 0 1 0 3
10. #M RVlP+i+PFS#ha#N #ha#M RVlP+PF##U RVlP 2
0 8 0 0
U La PF##M RVlP#ha#M RVlP+~# [Sl3]
1 1 0 0
ll. #!-RVlPiPFS#ha#N
8 #ha#RVlPPF##RVlP 2 +~-+La 3
PF##RVlP#ha#RVlP~# -[Sl5]

12. #!-RVlPiPFS#ha#N #ha#RVlPPF##RVlP

2 +~-+La
8 3
PF#et#RVlP#ha#RVlP~# -[Sl7]

An example of this, filling in roots and vowel patterns,

might be "toxniot ha radyo ha rgilot m~a'ammot et khal

ha som'im" ('The ordinary radio programs bore the listening

audience' ) .

Leading up to the analysis of roots and vowel patterns,

we list several classes of morphophonemes, which will also

be referred to later on.

If a and S are classes of elements, then aUS is the

class containing as members the elements of a and the elements

of S, i.e., it is the sum of a and S.

Sl8. Let G={', 6, X, h}

16
L={', t, n, y}

G+=GU{r}
L+=LV{l}l6

CR=G+UL+U{B, P, K, d, k, g, v, s, z, s,
c, m, N, Y , Y 2 , Y }
1 3
C=CRlJ{M, b, p, f, x, :}
-19-
 I.e., Vis the class of vowels with or without accent

(') or .. To indicate that I 'I

or 1
:
1
explicitly do not

occur, they are replaced by I vI

(giving 1
V
1 and 1 Vv 1 ,

respectively). Thus only v0 or v0 occur in env. __ ~, and

..._,
occur in env. When written over either

kind of bracket, then, just like the rest of the morpho-

phonemic environment, 1
' 1 and tvl hold for each element

in the bracket (e.g., { } or [ ], etc.).

W=CUV

~=the class of all morphemes (i.e., ·~~

is a variable ranging over morphemes.)

It will appear later that roots are discontinuous

morphemes. They are analyzed as follows:

319.

Numerical subscripts on C will refer henceforth to

position with respect to the root. Thus c0 will be the con-

sonant of a prefix, c1 will be the first consonant of a root

(or any consonant it.is transformed into), etc.

There are var~ous limitations on the distribution of

elements of CR in roots. I will not go into these here

(see pp.3-4 above) except for stating that N occurs only

as c1 , h rarely occurs as c3 , v is rare, and, when in R,

Yi occurs only as Ci (i=l, 2, 3) (only Y2 occurs outside

of R).

-20-
 Several subsets of R may be listed here for later
reference.

Ra=(sYr, sYm, ... } ' all of form CYC [see MRl]

Rb={kYm, sYB, ... } , 11 11 11
" [see MRl]
Rc={zYt, lYl, ... } , 11 11 11
" [see MR12]
Rd={smr, KtB, 'sP, ... } I f Cl=Yl, R£Rd [see MR2]
Re=R-Rd; i.e., all members of R except
those in Rd [see MR24]
I f Cl=y, then RERe

I f C =G, then RE:Re (except, sometimes, when

2
c2=X, e. g.' mXk, dXP)
If c 3=G, then RERe (unless c1 =Y 1 )
Re'={ysn, yr', ml', KBd, ... } [see MRl]
Re' is a subset of Re 1 7
Re 11 ={ 'mr, 'Bd, 'Kl, 'BY , 'PY }, all of the
3 3 form •cc. [see MR27]
Re" is a subset of Re
Rf= {KnP, 'nP, XBr, ... } [see MR26]
Rg={msl, Xkr, rkz, lcr, ... },mostly with
c =l,r,X. 1
[see MR24]

Vowel patterns are single or bi-vocalic discontinuous

sequences of elements in the range of V. The final vowel
of each is accented. The discontinuity will be represented
by '--', and the first and second vowels, respectively, or
any vowel into which they are transformed, by v and v 2 .
1
They are most conveniently treated as being each composed
of two elements, of which at least one is a vowel, and the
other a vowel or 0. Thus vowel patterns can be taken as of
the form •a 1 --B 2 1 , essentially, where a 1 , B2=V or 0.

-21-
 VlP 1 b={u--:o, u--:a, a--:u, a--:a:, i--:a, i--:o,
i--:u, e:--e, e:--a, e:--o, e:--u, o:--a, a--i,
a--u, a--o(:), a--e(: )}
where V2 --? V2 (:, when V2=o, i)

VlP 2={a--a, i--:e}

18
VlP=VlplaUV1PlbUVlP 2

There are also many four-consonant roots (and even

some with five consonants). This class is not treated in
the present study, but could conveniently be introduced
at this point. Four-consonant roots might be considered
to arise through replacement of : in VlP's of the form
v1 --:V 2 by a member of CR If a rule were given at this
point for the introduction of this fourth consonant in this
way, almost all such roots could be accommodated with al-
most no change in the subsequent statements of the grammar.
Various restrictions on distribution within M and U

can be stated now. In U,

VlP 2 ;t{~::&e1 in env. #{-~=}~-R- Ft [t:~

The non-zero members of M occur only with VlP's of the
0
form {~}--v. ml- predominates with v2=a, m2- predominates
with v2 =e, o. As an alternative treatment, we could dis-
pense with m2- and define two new classes of roots. This
change would involve only slight changes at various points.
Actually, a detailed statement should be given here of the
actual distribution of roots among vowel patterns. The

-22-
form of this statement might have some effect on the rest

of the grammar. Thus the system of eight verbal con-

jugations is considered in this treatment to be derived

from the two VlP's and two prefixes, each VlP occurring

alone, with each prefix, and with both prefixes at once.

Alternatively, this subsystem could be constructed from

two VlP's and three prefixes, each VlP occurring alone and

with each prefix. One of the considerations in rejecting

the latter formulation (though not the only one) involves

the statement (not given in this grammar) concerning the

distribution of roots through conjugations, which is greatly

simplified with the two prefix system. The simplification

of this statement is important because it makes clear the

active-passive relationship between conjugations. Further

statements on restriction of distribution will not be given,

but could be adjoined here.

At this point detailed statements of morphemic alter-

nation would also be given in a complete grammar. This is

simply a matter of added detail and would not affect the

formulation of the following statements. We might mention

the complex alternation of some roots of the form CYC:,

which sometimes become c1 c 3 c 3 , sometimes c1 Y2 c 3 , sometimes

Nc 1 c (:). These forms are rare, and details of the situ-

3
ation will be omitted. Another alternation characteristic

of several roots is the alternation YcC~NcC, the latter

appearing in env. Ft, or when U ~0.

0

-23-
3. Further Notational Statement

The statements of the grammar are in general of form

(1) and (2) (p. 6), but occasionally with certain minor

and generally self-explanatory deviations.

N9. If a statement is of the form

a--7 Sy in env. a __a , and y y where

1 1 2 1~ 2

then the y 1 referred to in the second half is that y

1
introduced by the first part, and not some other y in the
1
form under consideration (e.g., MlO, 11, 12). Furthermore,

the second part of a statement does not apply unless the

transformation indicated in the first part actually takes

place.

NlO. Statements of the form

a--7 [S~ y]

are abbreviations for

a~ 0 and B~ y

i.e., they assert that a is transformed into the change of

B to y.

Nll. Statements MR24, MR27, MR30, and MR32, under the

heading 'Intrusion of Vk' are to be interpreted as follows.

A substatement

a B

is an abbreviation for

-24-
and a statement

a S
y

is an abbreviation for

Nl2. Dots '· .. 1 are used to indicate sequences (per-

haps 0) where the particular shape occurring is inconse-

quential, or to indicate the relative position of two ele-

ments, in which case the element for which the dots stand is

placed above them in square brackets. Thus

.. ' [S]
a--, . env.
... y ln D
P

stands for

aS---t Sy (see MR7, MR22).

Nl3. No statement holds of a context including #,

unless this is explicitly stated. Thus statements hold

only within words.

Nl4. Juncture (non-phonemic) between morphemes is

written '+'. It is often omitted from statements. Thus any

statement holding of •aS• holds also of •a+S• (although the

converse is not true), and '+' can be freely dropped at the

end of the process of conversion to phonemes.

Nl5. If a statement is of form (2), where s1 =V,

B1 '=0, and y=CCa 2 , then it does not apply. I.e., 'V~ 0'
-25-
does not apply in env. CC. (Notice that :EC).

Nl6. u--7 u:, and similarly when u is introduced

below.

Nl7. When any v0 appears unmarked by

' .' or '-' in

a specification of relevant environment (i.e., in a or

1
y of a statement of form (2)), then the statement also holds

when v0 is replaced by any corresponding v0 , i.e., by v0 ,

v0 :, or v0 :.

4. Morphology and Morphophonemics

The sentences of the language are now represented in terms of

morphemes, some given in morphophonemic spelling. The fol-

lowing set of grammatical statements is designed to convert

any sequence of morphemes from section 2 into a sequence of

phonemes.

Below each statement several examples are given of

its operation. The references are to section 6, where a set

of sample morphological derivations is given.

MRl.
(l) m: (o)
~=J
(+La
1 2
, some-
times:-and
CYC ••• o:--- ul9

(2)
[~]: Ft, R{~~J
in
In u+ &2---7 env. #
'
( 3)
e
( 4)

-26-
E.g.

"naxon"
g-+BYn+a--a+La 1 --7g-+BYn+a--i:o+La 1 ; ultimately,
- -
"havinoti".
Also 6A4, 8.

(2) sYr+a--a+Ft+La 2 --7 sYr+a--i+Ft+La 2 ; ultimately,

"tasir"
kYm+a--a+Ft+La
2~ kYm+a--u+Ft+La 2 ; ultimately,
"takum".

(3) ~-+KtB+a--a+Ft+La
2~ ~-+KtB+a:--e+Ft+La~; ulti-
mately, "tikatev".
Also 6All.

(4) yr 1
+a--a+La
2~ yr'+a--e+La 2 ; ultimately,
"yare(y)ta"
ysn+a--a+!!!_----:. ysn+a--e+!!!_-; ultimately, "yasen"
KtB+a--a+!!!_----1 KtB+o:--e+!!!_-; ultimately, "kotev";
but kYm+a--a+!!!_--,l-7 kYm+o:--e+!!!_-; but ultimately, "kam".

MR2. (l) Y3 ... , __ ,, sometimes 1

[{c 3 :~~~~21}]
( 2)

( 3) [:+]h-+C1···
a:
' in
( 4)
env. m-
rLa(I) RE:Rd

( 5) 0
l- Ft t~:k ,
-27-
E.g.
(1) BnY+a--:a:~ BnY+a:--:a:; ultimately, "bana'i";
but zKY+a--:a:~ zKY+a--:a:; but is ultimately
"zakay" (see MR14).

(2) g-+Psk+a--a--7 g-+Psk+a--i:; ultimately, "hifsik".

( 3) g-+kdm+a--a~ g-+kdm+a--e; ultimately, "hekdem".

Also 6B5.

(4) 6A2, 6Al4.

(5) 6Al, 6A3, 6A5, 6A9, 6Al0;

smX+a--a+Ft+Lc~ smX+a--o+Ft+Lc; ultimately,
"lismoax".

MR3.

clc2c3{~-}+Ql--(:)Q2[~-J--7 clQlC2(:)Q2c3{~--}[~--J
and ::----1 :

E.g. each example in 6.

in env.

E.g. 6Al2, 6A9, 6All.

MRS.
La~~ li; a special case, henceforth, of 1. I.e., all
statements applying to 1 (from MR4) hold as well
of lr.
-28-
E.g. 6Al0.

MR6. and e--1 a

and u--7 o sometimes,
when c 2 tY 2 and c 1 tN or

(h)N in env. . .• ( +1)

E.g.

!'i_-+kiB:el+!!!_---7 kuB:al+!!!_-; ultimately, "mkubal"

!'i_-+KataB~ N+KataB; ultimately, "nixtav".
Also 6Al4, 6All.

MR7.

{n La1 _{_t}
!:£.--7

I
ra2](e)
~~~:; :~~~:
}
in env.

r
La
2

1
1_(+;E)

E.g. 6A7, 6A3, 6A5.

MRS.

E.g. 6A2, 6Al4; 6A6, 6Al0, 6All, 6Al2; 6Al, 6A3, 6A5, 6A6.

MR9.
V 1 --7{~}in env. ~··.

E.g. almost every example in 6A, 6B.

-29-
MRlO. n in env. • •• P

in env. _ ... cl
(l) La 1 ---7
[~_] i in env. [~_]
( 2) La 2 ---7 ta, and
a---+ 0 in env. _ j J
t---7 (e)x in env. s (a) (sometimes, but not
in env. P •. )

( 3) La ~ h, and
1

E.g. almost every example in 6.

MRll.

{~}-- 7 0 in env. #[~]+ ... _

E. g.'
'+samor+:E_~ '+samar; ultimately' "esmor"
n+KatoB+!:_--j n+KatoB; ultimately, "nixtov.

MR12.
0--7 {-~-}a{-(Y-3)} in env. C2_C {o:n#, Vl=i, sometimes}
3 !:_ (:E_, sometimes)

where RiRc

and if [~~=~ 3 ], v1=6, then V1 ---7 V 1 :[~~~etimes]

and Y2--7 y, sometimes
and aY c ---1 c a:Y , sometimes
3 3 3 3
-30-
E.g.,
ziKro:n---1 ziK:aron; ultimately, "zikaron";
but
k:ilso:n---1 (ultimately) "kilson".
Also 6B3, 6Bl;
kuPs+PF---1 kuPsaylF; ultimately, "kufsa'ot".
6B2, 6B7;
i:ons+~---1 &o:nas+~; ultimately, "'onasim"
k6sY (or k6saY ultimately,
3 +~~ k6:saY
3 +~ 3 +~);
"kosifm" ("ksa•fm").
Otherwise these last two would be "'anas:f.m", "ksa•im",
always.
taYs+~---1 tayas+~; ultimately, "tyasim";
but
BaYt+~---1BaYat+~; ultimately, "batim".

MR12'.
!'_~ (i )W(A, in env. except in env.{~~} __ )

E.g. 6Bl, 682, 6B3, 6B7, 6B8. Excluded: 6B6, 684, 6B5.
6Al, 6A4, 6A6, 6A7.

-31-
MR13.
(1) ;[+A

{:f~l} ... +_+~

0
C3+--22

mg}+ccaC+_
( 2) ~+A, exc.
sometimes, M c{~}c:{:}c+_ in
in env. env.
CCi:C+
+i

m{Q_-VCl ... ~-
y3 )
~--7 (3) F+A elsewhere in M
j
l
A (A)
(4) {+im+(A~--7 o:}t in M+, __A(#, ~~d t--7 0)
env.
except env. ~· ..

(5) +m+---7 +n+ in env. m

(6)

(7) env ·
t+l in
# { 1y+}
t+) 3
l
r) ... C (u) j(sometimes,l
and)
(and i---70,
(+u---7
23
and some- #na)
h } times)
(8) a in env. C __
{
3
(9) 0 in env. S ...

E. g.'

(1) m+roceY ultimately, "roca",

3 +~---7 m+roceY
3 +~A;
etc. The exception a~ ... is given to allow for the pos-
sibility (which may not be a real one) of forms like

-32-
"nivne(y)t", "mufne(y)t", alongside of "nivna", "mufna".
m+~-+Psi:k+E_ --7 m+~-+Psi:k+~A; ultimately,

m+N+KYo:n+~-7 m+N+KYo:n+~A; ultimately, "nxona";

6A2.
~+Y
1 sen+E_--7 !J:!.+Y 1 sen+E_A; ultimately, "ysena".
(2) m1 +ste:l+E_+£-7 m 1 +ste:l+~A£; ultimately,
"mastelat",
but
m +Plag+~+£-7
1 (ultimately) "mifleget";
similarly "xaverati" (from XaB~:r+F+S+La - - - 1 ) but "sxenti"
(from saK~:n+~+~+La 1 ); "mxona" (from m1 +KY6n+E_, ~-----1 ~A) but
"maxl6ket" (from m +Xl6:k+~);
1 "mi;tara" (from m +star+~,
1
~-7 E_A), but "magevet'' (from m +NgaB+E_); "kabala" (from
1
~ ~ ~ ~ ~ 24
kaB:a:l+~, ~--1 E_A), but "saxefet" (from saX:a:P+~);
"tipsa" (from tiP:es+~, ~---.:,~A), but "pikaxat" (from Pik:ex
+E_); "smira" (from sm:i:r+E_, E_--1E_A), but "gveret" (from
1
gB1:r+~
)
; 24 11
hadragat1"
1 (
from ~
~-+dra:g+~+l,
•
~---.:, E:_A ) ,
but "hakarti" as in "tat-hakarti" (from ~-+NKa:r+~+i);
"muxana" (from m+~-+KYa:n+E_, ~-7E_A), but 6Al4.
(3) Yald+E:_-7 Yald+E:_A; ultimately, "yalda";
6B4, 6B5. Exceptions are 6B4, 6B6, 6B7; "soxni'It" (from
m+so:Ken+u+~, ~--,L-1 ~A); "tizmoret" (from t+zmo:r+~, ~--H !:_A),
"bik;ret" (from Bik:o:r+E_, ~.. ,<4 ~A).
(4) 6Bl, 6B3, 6B8; 6A2, 6B4, 6B5; 6Al4, exception,
6B4, 6B6, 6B7.
(5) 6B4, 6A7.
( 6) y+KtoB+u+~-7 t+KtoB#na (ultimately, "tixtovna")
-33-
 or remains y+KtoB+u+~ (ultimately, "yixtvu);
y+KtoB+F_~ t+KtoB; ultimately, "tixtov".
I , I ' ,
(7) 1 +KtoB+u+~--7 1 +KtoB#na (ultimately, "ktovna")

or ---7 li+KtoB+u (ultimately, "kitvu")

I ' ------" I , . . '
1 +KtoB+~__, 1 +KtoB+1; ult1mately, "kitvi";

6A6.

(8) 6B7;
KataB+~---1 KataB+a; ultimately, "katva".

( 9) 6B6.

MR14. In M,

(2) ' in env.J

l~+
a j v{
... c 2a--
---
sometimes
}
(3) y in env. C

E. g.'
(1) "Yi:t:+o:n+a:Y #--1 'Yi:t:+o:n+a:Y i; ultimately,
3 3
"'itona•i"
Ban:a:Y
3
#--t Ban:a:Y 3i; ultimately, "bana'i"
laXY #---t leXi#; ultimately, "lexi"
3
k;:saY +im+A----t ko:sl+im+A; ultimately, "kosilm".
3
(2) Ban:a:Y
3
i--t Ban:a: •i; ultimately, "bana•i" (simi-
larly, 6 Yi:t:o:na:Y i, from (1))
3
kuPsaY~+o:tA--- kuPsa'o:tA; ultimately, "kufsa'ot"

kosaY imA--1 kosa'imA; ultimately, 11

ksa'im 11
3

-34-
 h-+rca: y 3+a+A--7h-+rca: I +aA; ultimate'ly' "hare a I a";
but
h-+Pla:Y +aA--7 (ultimately) "haflaya".
3
(3) h-+Pla:Y +aA--7 h-+Pla:y+aA; ultimately, "haflaya"
3
racu:Y --7 racu:y; ultimately, "racuy"
3
hav:a:Y ---) hav:a:y; ultimately, "havay"
3
'6nY +So---) 6 cmy+So; ultimately, "'onyo".
3

MR15.

::: [11}~ {~}:~:n:~v~+cVc 2]

in env.

"' ( '§._) #

E.g.'

(1) sadaY +o:tA--7 sad+o:tA; ultimately, "sadot'1 ;

3
6B3; 6B8;
racaY +onA---4 rac+o:nA; ultimately, "rac6n".
3
(2) malK+aA---1 malk+aA; ultimately, "malka".
Exceptions are zaKaY +ut, ultimately, "zxut" (not "zaxut");
3
ml+XnaY #, ultimately, "maxane"; malK+ut, ultimately,
3
"malxut"; malK#, ultimately, "melex".

E. g.'
m+N+r 1 aY #--7 n+r 1 aY ; ultimately, "nir 1 e
3 3
m+Ko:teB--7 KoteB; ultimately, "kotev"
-35-
6A2.

MR17.

E.g., 6B8;
m2+sKort--1 M+sKort; ultimately, "maskoret".

E.g., 6Al3; 6A4, 6A8, 6B5; 6Al4.

~-+N+zam:en--7 t+N+zam:en--7 Ntzam:en--7 Nztam:en--1
Nzdam:en; ultimately, "nizdamen".

MR19.
0~ a in env. #(l)Ck __c1 c 2a

i, Ckr! (or =)
u
and Ck=t
h1
a= e,
where a, and C-=h
k
o, and c;=Y 2
J
Ck=M

E. g. •

l+h+Psi:k--j lhaPsi:k; ultimately, "lhafsik"

t+Kli:t--1 taKli:t; ultimately, "taxlit"
t+KY 2 u:m--7taKY 2 u:m; ultimately, "takum"
t+rB+utA~ tarB+utA; ultimately, "tarbut"
-36-
 t+sBec~ tasBec; ultimately, "tasbec;
6B5;
m+kY 2 o: m--1 makY/>: m; ultimately, "makom"
N+KYon---1NaKYo:n; ultimately, "naxon"
M+sKort--tMasKort; ultimately, "maskoret".

MR20.

E.g., 6B6, 6B8; 6B7;

BiCy+o:t+A+~+xm--1 Bi6yo:t+eyASxm; ultimately,
"ba' ayoteyxem";
6B2;
siPr+~+nu--1 siPr+ey+~+nu; ultimately, "sifreynu".

MR21.
(1)
a{~( ... w1 )__ ((CV))#,
_ §.W, a=ey
0--7 A in env. [
(2)

E.g.'
(1) m+Xn+~+eha--7 m+Xn+~+eAha; ultimately, "maxaneha" (?);
6B4, 6B5; 6B2; excluded: 6B8,
slPr+ey+§_+nu-f7 slPrey~nu~.
siPrey~nu--7siPreyA~nu; ultimately, "sifr~ynu".
(2) Y1 arX+o:n--7 Y arX+o:n+A; ultimately, "yarxon":
1
6B6.

-37-
MR22.

in
env.
v1 _:{· · .C 3 : }and c:--7 c
sometimes

a.=0<A><S>
J -

E. g.'
(l) 6Al3;
ht+kaY 2 :~m+ut--1 ht+ko:m~m+ut; ultimately,
"hitkommut".
(2) ht+kaY 2 :em----1 ht+kay:em; ultimately, "hitkayem"
BaY 2 t#--1 Bayit#; ultimately, "bayit";
6B6;
zaY 2 t+imA--7 zayt+imA; ultimately, "zeytlm".

( 3) 6A2, 6A4, 6A8, 6B7;

siPar+aY
2 +A~+o--7 siPar+av+A~; ultimately,
"sfarav";
6B8; 6Al4, 6B6.

MR23.
(l)

Vi~ 0 in env.
( 2)

E.g.'
(l) daBar+imA~ dBar+imA; ultimately, "dvar:f.m";
-38-
6B2,

daBar+ey+A+~+A--7 dBar+ey+A+~+A--7 dBr+ey+A+~+A;

ultimately, 11
divrey 11
;

6B4, 6B7.
(2) Bay:as+a:n~Bay:s+a:n; ultimately, 11
baysan".

MR24. Intr1.1sion of e
( 1) c__ •__ck, ~~3, V=i:,e (only sometimes whenl
(V) - in env. L' __ )26
27

(2) In u+ lr~L) sometimes

}

( 3) L' c~c
yl

(4) m__CC(a)C, where CCC£Rg

2
(5) #G~~ c1 C__ )C~ (where i=l, c =G) 7
1
<i>

unless: G.=h in U+ 28
1

( 6)

E.g.,

(1) 6A8,

mo:ce't~mo:c~•et; ultimately, "moce(y)t";

6A6; 6Al2; but sometimes y+'r~z~ ye'er~z, but is ultimately

"ya•ar~z".
(2) h+'di:P---t he,edi:P; ultimately, "he'edif"

N+'zaB---t Ne'ezaB; ultimately, "ne'ezav";

excluded is N~,s~Y
3; ultimately, "na'asa" or "ne'esa"(?);

-39-
 t+XKam----1 teXKam; ultimately, "texkam".
( 3) 6A5, 6A9, 6Al0.

( 4) m+rKaz--1 merKaz; ultimately, "merkaz".

( 5) '+"soY 3 __,
-, •e6esoY 3'· ultimately ' "'e'ese" '
sometimes (ultimately "'a'ase" if (5) does
not operate);
6B3; '+smor--7 'esmor; excluded are h+sPi:k, ht+mas:er+
u:tA, etc., also '+'s6Y 3 sometimes (as above), and '+N+
Y1 a:de', ultimately "'ivad&'", sometimes, instead of
"'evada'"·
(6) 1 I +haki' :m----1 1I +hakem;
- -
ultimately, "hakem"
taPsi:k#na--7 taPsek#na; ultimately, "tafsekna".
These are essentially literary forms.

MR25.

(1) {;y}in env.{#~l~·~}--[~J

( 2) t in env. U, _a#
( 3) 0 in env. w
c4) a. in env. U, #C....
J_ -
, where Cj__"'{~}, j__:O'

or where j__=l
C5) e
E.g.,

(l) racaY +ti----1 raci+ti

3
Xik:eY +ti~ Xik:i+ti; ultimately, "xikiti"
3
hsvi:Y +nu----1 hsvey+nu; ultimately, "hisve(y)nu".
3
(2) racay +a~ ract+a; ultimately, "racta".
3
(3) 6Al2;

-40-
 zKY zKu:yo:tA; ultimately, "zxuyot".
3 u:yo:tA~
(4) hrci:Y __, hrca; ultimately, "hirca"
3
racaY ____, raca.
3
(5) mXzaY __,. mXze; ultimately, "maxaze"
3
yrcoY ~ yrce; ultimately, "yirce".
3
MR26.

0~ {~}in env. #{(i)}c __cc{~~~e:~~~ 1}

E. g.'
KnP+eyASA__,. KanP+eyASA; ultimately, "kanf~y".
6Al, 6A3, 6All, 6Al3, 6B4, 6B8;
t+ysan-4· tiysan~ tiSan.

r
MR27. Intrusion of o
(1) 6a c27

( 2) (- ) 3c22 )
u:- G+:, sometimes
(3) L c2 , where Re:Re"
e1e
(4)
VY 1

E. g.'
(1) c6hr+a:yAim--;.. cohora:yAim; ultimately, "cohorayim".
(2) Xu:k:--;.. Xok:; ultimately, "xok"
yaXu:g:u--;. yaX6g:u; ultimately, "yaxogu"
mPu: ras-:;.. mPoras; ultimately, "mforas"
But
mY 1 U: 1
: as~ (ultimately) 11
myu 1 as",

-41-
 (3) 6A6.

(4) yaY 1 r:i:d~ yori:d; ultimately, "yorld".

MR28.

<~> in (l){U+
vo~ 0 env. c SWJj - - -C.=N
- )
)-- 3[ - - - -
( 2) M,
WCiCi V Vo•e,ic~-,}'[ellsewhere,]
J... rarely

E. g.'
(1) 6Al, 6A7, 6A8;

KoteB+imA--;;. .KotB+imA; ultimately, "kotvim"

histo:vev+u:tA~ histo:vvu:tA; ultimately,

"histovv~t".
( 2) miNk~l+S+oA~ miNkl+SoA; ultimately, "makl~"
tiP:~s+u:tA--;. tiP: s+u: tA; ultimately, "tipsut"

mizBeX+o:tA--;. mizBX+o:tA; ultimately, "mizbxot"

but
~

maklet+imA--;. (ultimately) "makletim".

MR29.
v l·n env. {tN- , sometimes}}
yl~
{
Y

E.g.'
hitY 1 aK:eX~ hitvaK:~X; ultimately, "hitvak~ax"

but

hitY
1 as:eB~ hityas:eB; ultimately, "hityasev".

6All, 6B6.

-42-
MR30. Intrusion of a

and

(2)
mi 1NC 2
x(_)c ••• ,
2 where ... does not contain c 3 (but)
does not hold sometimes in env .
. . . S)

( 3) #G C

(4) 1+
-
c2
e
(5) In u, __
~ c 3 [~J
v2

<~>l
l {
~ ---} c 27, where ... contains noW,
WJ... ••• G <:> ... - ~
Vj_=i, e 2
G=C only in M+ or
3
1+... 29

E.g.'

(1) noBeX--;. no:BeaX; ultimately, "noveax"

likroG---l> likroa 6 ; ultimately, "likroa'".
But

naBaX-A> naBaaX
likro' -,4 likroa'; but is ultimately, "likro'".

-43-
6Al0; 6All;
sBe6SA (as in "sva' racon")~ sBea6e_A~ sBa6e_A;
ultimately, "sva'"

mixBeaX~A~ mizBeaXSA~ mizB~X~A; ultimately,

"mizbax", sometimes.
gaBoh~ gaBoah~ gaBoha; ultimately, "gavoha"
goBh~ goBah---':l> goBa; ultimately, "gova"
gBohe_A~ gBoah--7 gBah--7 gBa; ultimately,
"gva" (as in "gva koma").
(2) miNBat~ maNBat; ultimately, "mabat"
miXze~ maXaze; ultimately, "maxaze";
6B8.
(3) 6A4; 6A5; 6A8;
XBer+imA---:;. XaBer+imA; ultimately, "xaverim".
( 4) 6A9, 6Al0.
(5) 6A4, 6Al3.
(6) 6B4;
PiXd~ PaXad; ultimately, "paxad"
'o:reX+t~ 'o:raXat; ultimately, "'oraxat"
lade•+t~ lada 6at; ultimately, "lada'at";
6A7, 6Al4;
roXB--;. roXaB; ultimately, "roxav".
But
karaGta-,4 kara6ata; but is ultimately "kara'ta".

MR31.
e.CA)~ 0 (in env. __#)

E.g., 6B2, 6B4, 6B5, 6B6, 6B7, 6B8.

-44-
MR32. Intrusion of e

c. l3o
(1)
rr:T
l
(') G~:,
+
sometimes
+
NG , G =C
1

c j_r.
~ .. Vk=i<:>,a
'

a<:>
(4) c
'e
( 5) In u+, #{h} c V a=i a{except in.env. __ ...
ma 1 ' ' or' sometlmes' a=0
o} and e~

(6) __w, in certain dialects 32

ey

E.g.,

(1) ki:c:_..;;.. kec:; ultimately, 11

kec"
higi:n:--;.. higen:; ultimately, "hegen"
'iX:aru--;.. 'eX:aru; ultimately, "'exaru";
but
riX:amu_..;;.. (ultimately) "rixamu"
yiN6"ase_..;;.. yeNb"ase; ultimately, "ye'ase".

KoteBt_..;;.. KoteBet; ultimately, "kotevet";

6A9; 6B4;
-
gBi:rt~ gBi:ret--;.. gBeret; ultimately, "gveret".

-45-
But
Pi: 1-,L:.. Pi: el.

(3) BaytiA~ BeytiA; ultimately, "be(y)ti";

but
Bay:sa:nA~ Bey:sa:nA; but is ultimately "baysan".
(4) moc'et~ moceyt; ultimately, "moc~(y)t;
6A8.
( 5) 6A4, 6A8. But
haBi: n;ti--7 (ultimately) "havinoti"
(also hiti.:B--7 heyti.:B; ultimately, "he(y)ti.v".)
mBi:nimA--7 meyBi:nimA; ultimately, "me(y)vinim",
sometimes.

MR33.

r~v}-. 0

E.g., 6Al0;
lakaXt!'_--7 lakaXt; ultimately, "lakaxt";
6A7, 6Al3.

MR34.

-46-
E. g.'

kiB:~l---7 kib:el; ultimately, "kibel"

BaXu:r~ baXu:r; ultimately, "bax~r"
hitBar:~r~ hitbar:~r; ultimately, "hitbarer";
6A3, 6B5.
yiNtBar:~r--7 yiNtbar:er; ultimately, "yitbar~r".
Many examples in section 6.

l
MR35.

- - - } 10 in env. • •• A
{(~~- •• A~ V, where ... contains noV

E.g., many examples in 6B; 6A2.

MR36.
h~ 0 in env. C. , c.'/y,:
J...- J...

E.g., 6B5;
sifrha~ slfra.
But
"" ""
sifreyhem-f.7 sifreyem
,. ..
'o:h~v~ o:ev.

MR37.
V-7 V in env. . .. V

E.g., 6A2, 6B2, 6B5, 6B6.

MR38.
'-'

in env. _<:>C:W, C'/G +

{., ,;metime,}-"'
-47-
E. g.'

xok:im---7 xuk:im; ultimately, "xukim"

~

kec: im---7 kic: im; ultimately, "kicim"

ma:s:im---7 mis:im; ultimately, "misim",

but

ra:b::i.m-17 rib:im; but is ultimately "rabim".

But

mfo:':ar--Jt mf'u'ar; but is ultimately "mfo'ar".

MR39.

#--? + in env. na#

E.g., 6A6.

MR40.

N-7 [in env. c.v J._-_ ck'

E.g., 6A3, 6All. But

liNpol--? linpol
~

hiNXil--7 hinXil; ultimately, "hinxil".

MR41.
X-? X

E.g., 6Al4, 6B3, 6B8.

MR42.

E.g., 6A7, 6All, 6Al0, 6B4, 6B5.

-48-
MR43.

M~ m

E. g.'

Maskoret--7 maskoret.

MR44.

W:--7 W

E.g., 6A2, 6All, 6Al3, 6Al4, 6Bl, 6B3, 6B5, 6B8.

MR45. If there is noV in env. # ... __ ... # 2 , then

1

E.g., 6Al, 6A3, 6A5, 6A7, 6Al0, 6All, 6Al2.

This completes the morphological and morphophonemic
section. Running once through the whole set of transfor-
mation statements given above, both syntactic and morpho-
logical, gives one sentences in phonemic representation. 33
Running through the whole set in all possible ways, assuming
the gaps cited to be filled in, would give all possible
sentences. A phonemic statement giving allophones of phonemes
(and describing automatic intrusions, e.g., intrusion of
schwa in env. CC __C, etc.) would complete the grammar.

5. Justification

The fundamental question about any grammar, aside from

that of its adequacy in describing the facts, is the question:
why is it constructed in the particular way it is. As

-49-
stated in the introduction, we will attempt to give a limited
answer to this question for the morphological statement
just presented by demonstrating that the statements are
partially ordered by criteria of simplicity. The necessary
ordering of the statements of the morphology is given by
the following chart. In this chart, a line drawn from
left to right, not crossing horizontal lines, determines
a necessary order (left to right) among the elements
in the boxes through which it passes. These elements are
the numbers of the morphological statements given in the
preceding section. Any two elements through which such
a line passes have an order defined for them. For each
statement, it can be seen at a glance which statement it
must precede, and which it must follow.
Within the statements themselves, substatements are
ordered, in general. Indeed, even within a bracket, each
line generally must precede each succeeding line. No proof
of this will be given for parts of numbered statements,
but this could easily be done in a manner similar to that
exhibited below for the main statements.

12 14 15 29
7 25 2E 27 1.!5
10 l l 12' l
3 4 20 'i6
22 P3 24 31 32 35 41
1- : 30 37,3 11
l - 6 8 13 17 25 '.. 44
r----' 33 34 l4ol41
2 16 f-- E 28 39
5 9 18 42
43

-50-
 The ordering indicated by the broken line (MR25<MR28)
is that given by the application of criterion 2 (p. 52) as
a subsidiary criterion.
To justify the ordering given by the chart (in a
limited sense--seep. 4), then, it is necessary to justify
the construction of each vertical line segment in the chart.
One way to do this will be to show that if the statements
separated by this line segment were applied in reverse
order, the wrong form would result. In such cases the jus-
tification of the ordering MRm<MRn will be given in the
following form:

n
MR!!!<MR:Q_, otherwise a-=::> a'---7 *a" [a''']

This means: MR!!! precedes MR:Q_, otherwise some sequence a

will be transformed by MR:Q. (before MR!!! is applied) into
a sequence a', ultimately resulting in a", which is in-
correct, the correct form being a'''·
Alternatively, we may justify the ordering MR~<MR£

by showing that, were they to be interchanged, they would

have to be complexly rephrased to generate the same forms.
The justification of the ordering MRm<MRn will in this
case be of the following form:

MR~<MR£, otherwise a'/a in MR!!! (or in MR~)

meaning that MR!!! precedes MR~, otherwise MR!!! (or MR~)

would have to be rephrased with a' replacing a, a' being

less simple.
The criteria for justification of ordering are as

-51-
given at the conclusion of section 0: simplicity is in-

creased by

1. reduction of the number of symbols in a state-

ment (paired brackets, etc., counting as one

symbol);

2. reduction of the length of derivations,

with the second requirement subsidiary. Actually it applies

only once, and then in a trivial fashion. I mention it

only to indicate explicitly that this consideration, taken

as subsidiary, will not materially increase the ordering

restrictions.

Once the criteria are accepted, it is necessary to apply

them rigorously. Thus some of the justifications are based

on trivial considerations. A higher order justification,

demonstrating the ordering imposed on statements two,

three, etc., removed, while it would be more complicated

to present, would show many less trivial restrictions.

That is, if the reason for MRm<MRn is apparently trivial,

it is almost always the case that the reason for MR~<MR~+~,

and form MR~-~<MR~ (where ~. ~are small integers, generally

1) is not trivial.

For justification of the first form, when it is not

directly stated, it is implied and can be shown that if there

is an effective possible reformulation, it is more complex.

MRl<MR2, otherwise
~ MR2 ~
£-+BYn+a--a+La 1--7 £-+BYn+a--i:+La 1 --? * always,

-52-
 "hevanti" ["hevanti" or "havinoti"].

MRl<MR3, otherwise
( [C3~] , in MRl.
...... )/(o)

MR3<MR4, otherwise

MR2<MR5, otherwise
1
1, lfl /Lc in MR2.

MR4<MR6, otherwise

({~:~})/(+1) in MR6.
MR5<MR6, otherwise

c{::i})/(+1) in MR6.
MR6<MR8, otherwise

This more than compensates for the fact that

... (+1)/(1) when MR6<MR8 .

MR7<MR10, otherwise
La !=2_~
10 t!£~ -
*"at" -
["ata"].
2

MR8<MR10, otherwise MR8 must list all forms into which

MRlO transforms Lak.

-53-
MR8<MR9, otherwise "or in env. -···{~"must be added to

MR9.

MRlO<MRll, otherwise
~ - ll La +samor----;;
La +samor+£--;. ~ - ~ '
*"'esmor" ~ -
["nismor"].
1 1

MR12<MR12', otherwise

im+A/£ in MR12.

MRll<MRl2', otherwise

... c ; ... in MRll.

3
Also, criterion 2 is violated, since

MR12'<MR13, otherwise the conditions for £~ imA would

have to be added to MR13(4), and ".£--7 0 in

env. o:t" would have to be added to MR12'.

MR9<MR13, otherwise

MR13<MR14, otherwise

{~;v in MR14(2).

MR13<MR20, otherwise the environment for ~eY

2 would be
very complicated to state, in MR20. And

Y
- . 20 -
1 alad+lmA~+~o--? Y
1 alad+imA~+~a:Y 2 +~o--7

*"ylad~tav" ["yaldotav"].

MR13<MR16, otherwise

m+kY 2 a:m+~~ kY 2 a:m+~--7 *"kamet" ["kama"].

-54-
MR14<MR15, otherwise
~ 15 ~ ~ J
b_-+rca:Y +aA---:;. g-+rc+aA---7 *"harca" ["harca'a" ,
3
6 ani:Y +imA~
3 6an+imA---:;. *"'anim" ["'aniyim"].

MR16<MR17, otherwise "in U" must be added to MR16.

MR16<MR18, otherwise CY a/CY in MR16.

2 2
MR15<MR21, otherwise ~A/~ in MR15.

MR20<MR21, otherwise both have to be completely and com-

plexly rephrased, with many repetitions, since
A is introduced by MR21 on the basis of the
transformations resulting from MR20, which, if
the order were inverted, would then have to be
listed twice.

MR17<MR19, otherwise

Jck };c in MR19, and several consequent

lmi Ci=l, 2 ) k complications.

MR18<MR19, otherwise
~ 19
t+b_-+Psi:k---:;. t+b_-+Psi:k--7 *"tifsik" ["tafsik"].

MR2l<MR22, otherwise "or ... ", where ... is a list of all

forms into which a:Y is transformed by MR22,
2
must be added to MR21(2).
~ 22 ~
siPar+aY
2 +A+~+x--7 siPar+ayA§.ix--7 *"sfarayix"
[ "sfarayix" J.

MR19<MR22, otherwise

-55-
 mkY 2o:m} *j"mkom"}{["makom"]}
tkY 2u:m ---.:,. "tkum" ["takum"]
{
etc. etc. etc.

MR19<MR43, otherwise
' 43 m+smer---.:,.
M+smer--7 ' '
* "mismer" '
[ "masmer"].

MR22<MR23, otherwise
' 23 maKY o:naA._....::,..
maKY 2 o:naA--7 ' '
*"maxona" '
["mxona"].
2

MR23<MR24, otherwise
' 24 '
6igalo:t§.A---7 b"igalo:t§.A~ *"'iglot" ["'eglot"].

MR24<MR25, otherwise all forms into which Y is changed by

3
MR25 must be specified in MR24(2).

MR24<MR28, otherwise
Co)/o in MR24(3).

MR25<MR26, otherwise
26
zKY uyo:tA--7 ziKY *"zixuy6t" ["zxuyot"].
3 3 uyo:tA~
MR25<MR28, otherwise criterion 2 would be violated since
' 28 25
ra'aY +a--;:;. ra'Y +a--7 ra't+a; ultimately, "ra'ata",
3 3
instead of simply
' +a--7
ra'aY 25 ra'ta; ultimately, "ra'ata".'
3

MR26<MR27, otherwise
#C _ _! _ in MR27(4).
(V)Yl VY1

MR27<MR29, otherwise "except in env. V " must be added to

MR29.
-56-
MR27<MR30, otherwise
- 30
cohrayAim--7 coharayAim--7 *"coharayim" ["co-
horayim" J.

MR28<MR30, otherwise
- 30 Ka'aBu-7
Ka'aBu----7 - - ["ka'avu"].
*"ka'vu" -

MR30<MR31, otherwise
31 miXno:tA-7 *always "maxanot"
miXno:tA!2_A--7 -

[ "maxan6t" or "maxn6t"].

MR30<MR39, otherwise
- 39 tiKtoB+na----7 *"tixtavna" ["tixtovna"].
tiKtoB#na--7 -

MR30<MR42, otherwise
'/42 zro'----?
zro"'--7 - -
*"zro'" -
["zroa'"].

MR3l<MR32, otherwise
<!2_><A>/<A> in MR32(2).

MR32<MR35, otherwise
- 35 siPrxm--7 *"sifrxem"
siPrxmA-7 - -
["sifrxem"].

MR32<MR33, otherwise
I - 33 - - ,
1 +hakem--7 hakem---:;. * "hekem" [ "hakem"],

KataBt~4 KataBt--7 *"kat a vet" [ "katavt" J.

MR35<MR45, otherwise
45 sfareAxa----:;. *"sfarexa"
sf'areAxa--;,. - ["sfarexa"].
-

MR35<MR36, otherwise

-57-
MR35<MR37, otherwise "or in env. ... A" must be added to
MR37.

MR33<MR34, otherwise
1
1 iKtoBl.! 1 1 ixtov__,.. *"xtov" [ 11 ktov"J.

MR37<MR38, otherwise
- -
Xok:im~ 38 Xok:im___;;;.
- - - 11
* 11 xokim [
11 - 11 ] .
xukim

MR38<MR4l, otherwise
C~G+, C~x/C~G+ in MR38.

MR38<MR44, otherwise
- 44 Xokim___;;;.
Xok: im___;;;. - -
* 11 xokim* - 11 ] .
[ 11 xukim

MR34<MR44, otherwise
kiB:el~
- 44 kiBel----7
- - 11
* 11 kivel [
11 - 11 ] .
kibel

MR34<MR40, otherwise
-
maNBat~ 40 -
maBat~ * 11 - 11
mavat [
11 - 11 ] .
mabat

MR40<MR4l, otherwise
G or x/G in MR40.

6. Sample Derivations

Several instances will be given here of the application of

the morphophonemic and morphological statements of section
4, in the same manner as the examples (on pp.l8, 19) of
the application of the syntactic transformations. In each
example, the heading is a sequence of morphemes, the final
line the corresponding sequence of phonemes, and the number
-58-
of each transformation is again placed in parentheses to
the right of the sequence resulting from the application of
the transformation. Section A consists of selections
from u+, thus, essentially, verbs; section B consists of
selections from M, thus, essentially, nouns (see p.l4).

A.
l. KtB+a--a+Ft+Lal+~ 2. kY 2 m+a--a+~-+~
l. KtB+a--o+Ft+Lal+~ [MR2] l . kY
2 m+a--a:+~-+:E [MR2]
2. KatoB+Ft+La +!:_ [MR3] 2. kaY 2 a:m+~-+:E [MR3]
1
3. Lal+Kat6B+!:_ [MR8] 3. m+kaY
2 a:m+~ [MR8]
4. Lal+KtoB+!:_ [MR9] 4. m+kY 2a:m+;E [MR9]
5. y+KtoB+~ [MRlO] 5. m+kY
2 a:m+~+A [MRl3 .1]
6. y+KtoB+u [MR12'] 6. m+kY 2a:m+a+A [MR13.4]
7. yiKtoBu [MR26] 7. kY 2a:m+aA [MR16]
8. yiKtBu [MR28] 8. ka:maA [MR22]
9. yixtvu [MR34] 9. ka:ma [MR35]
10. yixtvu [MR45] 10. ka:ma [MR37]
[MR44]
~

ll. kama

l.:.. Lal+~#NPl+a--a+Ft+Lal 4. g-+kY

2 m+a--a+Lal+~
~

l. La3Lp#NPl+a--;+Ft+La3 l. g-+kY
2 m+a--i:+Lal+~
[MR2] [MRl]
~

2. La3Lp#NaPol+Ft+La3 2. g-+kaY 2i:m+Lal+!:_ [MR3]

[MR3]
3. La3u#NaP6l+Ft+La3 [MR7] 3. g-+kY 2i:m+Lal+!:_ [MR9]
4. La3u#La3+NaPol [MR8] 4. _Q_-+kY 2i:m+n+!:_ [MRlO]
5. La3u#La3+NPol [MR9] s. g-+kY 2 i:m+n+u [MR12 1 ]

-59-
3. (cont.) 4. (cont.)
6. hu#yNPol [MRlO] 6. h+kY 2i:m+nu [MR18]
7. hu#yiNPol [MR26] 7. h+ki:mnu [MR22]
8. hu#yiNpol [MR34] 8. haki:mnu [MR30.3]
9. hu#yipol [MR40] 9. hakamnu [MR30.5]
10. hu#yipol [MR45] 10. he(y)kamnu [MR32]

~ La2+~#Y
1 c•+a--a+Ft+La2 6. 'Kl+a--a+Ft+La2+~+~

l. La2+~#Y
1 c•+a--o+Ft+La2 l. 'aKal+Ft+La2+P+F
----- [MR3]
[MR2] 2. -La2+'aKal+P+F
-- [MRS]
2. La2+~#Y 1 aco'+Ft+La2 3. La2+'Kal+P+F [MR9]
[MR3] 4. t+'Kal+PF [MR10]
3. '+La2#Y 1 aco'+Ft+La2 5. t+'Ka1+uF [MR12']
[MR7] 6. t+ 'Kal#na [MR13.7]
4. 'La2#La2+Y 1 aco' [MR8] 7. te'eKal#na [MR24]
5. 'La2#La2+Y 1 co' [MR9] 8. toKal#na [MR27]
6. 'ta#t+Y 1 co' [MRlO] 9. toxal#na [MR34]
7. • ta#tece' [MR24] 10. toxalna [MR39]
8. 'ata#tece' [MR30]
9. 'ata#tece' [MR45]

L.. La3+~+PF#r'B+a--a+La3+PF ~ g-+BY 2 •+a--a+Lal

l. La3LpPF#ra'aB+La3PF l. ~-+BY
2
1
+a--i:+Lal [MRl]
[MR3] 2. g-+BaY 2:£: '+Lal [MR3]
~
r,~
2. La3ePF#ra aB+La3PF [MR7] 3. g-+BY 2i: '+Lal [MR9]
3. hePF#ra.;aB+PF [MRlO] 4. g-+BY 2 i: '+ti [MRlO]
(~
4. hemF#ra aB+u£:_ [MR12'] 5. h+BY 21: I +ti [MR18]

-60-
7. (cont.) 8. (cont.)
5. hen#ra'aBuF [MR13] 6. h+Bi: I +ti [MR22]
6. hen#ra{BuF [MR28] 7. h+Be'eti [MR24.1]
7. hen#ra?aBuF [MR30] 8. h+B'eti [MR28]
8. hen#ra'aBu [MR33] 9. haB'eti [MR30]
9. hen#ra'avu [MR34] 10. haBeyti [MR32.4]
10. hen#ra'avu [MR42] 11. heyBeyti [MR32.5]
11. hen#ra'avu [MR45] 12. he(y)Be(y)ti [MR32.6]
13. he(y)ve(y)ti [MR34]

C ~ I
~ Y1 sB+a--a+Ft+Lc 10. Y1 d +a--a+Ft+La 2

Y1 sB+a--.;+Ft+Lc ? , I
l. [JVIR2] l. Y1 d +a--o+Ft+La 2 [MR2]
2. Y1 ado-c +Ft+La r2
v'
2. Y1 asoB+Ft+Lc [MR3] [MR3]
3. Y1 asoB+Ft+Ht [MR4] 1
3. Y1 actoc+Ft+1 [MR5]
4. HY 1 asoB+t [MR8] 4. 1 I +Y 1 ado, c [MRS]
5. l+Y 1 soB+t [MR9] 5. 1I+Ycta"' [MR9]
6. 1eseBt [MR24] 6. 1Iedeb [MR24]
7. 1as~Bt [MR30] 1 [MR30.1]
7. 1 eda 6
8. 1aseBet [MR32] 8. l 1 ada 6
--
9. lasevet [MR34]
9. da•
[MR30.4]
[MR33]
10. da' [MR42]
11. da' [MR45]

11. ~-+Y 1 d'+a--a+Ft+Lc 12. 1 PY 3+a--a+Ft+Lc

--

l . N-+Y d '+a: --e+Ft+Lc l. 'aPaY +Ft+Lc [MR3]

1 3
[MR1] 2. 'aPaY 3+Ft+Hot [MR4]
~c
2. ~-+Y +Ft+Lc [MR3] 3. 1+'aPaY 3+ot [MR6]
1 a:do

-61-
11. (cont.) 12. (cont.)
3. :t!_-+Y 1 a:doG+Ft+l [MR4] 4. 1'PaY ot
3 [MR9]
4. hN+Y 1 a:do'+Ft+1 [MR6] 5. 1e'ePaY ot [MR24]
3
5. l+hN+Y 1 a:do" [MR8] 6. 1e'ePot [MR25]
6. 1hiNY 1a:do' [MR26] 7. 1e'efot [MR34]
7. 1hiNva:d6b' [MR29] 8. 1e 'efot [MR45]
8. 1hiNva:da' [MR30]
9. 1hiva:d8J [MR40]
10. 1hiva:da' [MR42]
11. 1hivada' [MR44]
12. 1hivada • [MR45]

13. £-+sY 2 B:+i--:~+La2+~ 14. ~-+!l_-+NY 2 X+a--a+~+~

1. £-+siY
2 :eB+La2+~ [MR3] l.
-h-+N-+NY
- 2X+a--a:+m-+F
- -
2. £-+saY 2 :eB+La2+~ [MR9] [MR2]
3. £-+saY 2 : ~B+t+~ [MR10] 2. £-+!l_-+NaY
2 a:X+~-+~

4. hstaY 2 :eB:t~ [MR18] [MR3]

5. hsto:B~BtF [MR22] 3. £-+u+NaY [MR6]
2 a:x+~-+~
6. histo:BeBtF [MR26] 4. m+~-+u+NaY 2 a:X+~ [MRS]
7. histo:BaBtF [MR30] 5. m+~-+u+NY
2 a:X+~ [MR9]
8. histo:BaBt [MR33] 6. m+~-+u+NY 2 a: X+t
9. histo:vavt [MR34] [MR12.4]
10. histovavt [MR44] 7. muNY 2 a:Xt [MR18]
~

8. muNa:Xt [MR22]
9. muNa:Xat [MR30]
10. muna:Xat [MR40]
11. muna:xat [MR4l]
12. munaxat [MR44]

-62-
B.

1. mlK+a--0+!:_+~ 2. mlK+a--0+!:_+~

1. malKPF [MR3] l. malK+ !:_+~ [MR3]

2. malaKPF [MR12] 2. malaK+P+S [MR12]
3. malaKimAF [MR12'] 3. ma1aK+imA+S [MR12']
,
4. malaKo:tA [MR13] 4. malaKeyA~ [MR20]
5. mlaKo:tA [MR23] 5. malaKeyA~A [MR21]
6. mlaxo:tA [MR34] 6. malKeyA~A [MR23]
,
7. mlax~:t [MR35] 7. malKeyA [MR31]
8. mlaxot [MR44] 8. malxeyA [MR34]
, ,
9. malxey [MR35]
10. malxey [MR37]

.h XBr+ i--0+ PF 4. c'k+a--a+F+S+La2+P+F

-----
,
1. XiBr+P+F [MR3] 1. ca~ak+F+S+La2+P+F
- - - - - [MR3]
,
2. [MR12] 2. caCak+FS+x+PF [MRlO]
XiBaY
3 r+!:_+~ - -
3. XiBaY [MR12'] 3. ca'ak+FS+x+mF [MR12']
, 3 r+im+A+~ - -
4. XiBaY r+o:tA [MR13] 4. ca '"ak+at~+x+m~
3
s. XiBro:tA [MR15] [MR13.3,4]
6. XeBro:tA [MR24] 5. ca 6 ak+atS+xn [MR13.5]
7. Xevro:tA [MR34] 6. ca'ak+atS+xnA [MR21]
8. Xevro:t [MR35] 7. c'akatsxnA [MR23]
9. xevro:t [MR4l] c'katSxnA [MR23]
10. xevrot [MR44] 8. ci'katSxnA [MR26]
9. ca 'akat~xnA [MR30]
10. ca 6akatxnA [MR31]
11. cacakatxenA [MR32.2]

-63-
 4. (cant.)
12. cab"akatxen [MR35]
13. ca'akatxen [MR42]

9c+sr'+a--a+~+~+La3+~ 6. Y
~ 1Y2 d+a--a+a:Y
2 +~+~+La2+!
l. h-+sPb+a--a:+~+~+La3+~ l. Y
1 aY 2 ad+a:Y 2 +~+~+La2+~
[MR2] [MR3]
2. h-+saPa:b+~+~+La3+~ 2. Y 1 aY 2 ad+a:Y 2 +~+~+x+~
[MR3] [MR10]
3. h-+sPa:'+~+~+La3+~ [MR9] 3. Y 1 aY 2 ad+a:Y 2 +m+~+x+~
4. h-+sP~ J +F+S+h+P [MR10] [MRJ:2']
5. h-+sPa:(+F+S+h+m [MR12'] 4. Y
1 aY 2
'
ad+a:Y
2 +m+~+x
6. h-+sPa:'+at+S+hm [MR13.9]
[MR13.3,4] 5. Y
1 aY
2 ad+a:Y
2 +aY
2 +~+x

7. h+sPa:'+at+S+hm [MR18] [MR20]

8. hasPa:'+at+S+hm [MR19] 6. Y
1 aY
2 ad+a:Y
2 A+aY 2 +~+x
9. hasP~:?+at+S+hmA [MR21] [MR2l. 2]
10. hasPa:'+at+S+hamA '
7. Y
1 aY
2 ad+a:Y 2 A+ayi~x
[MR30.6] [MR22.2]
11. hasP aJ +at+hamA [MR31] 8. Y
1 ad+aAyi~x [MR22.3]
12. haspa J +at+hamA [MR34] 9. yadaAyi~x [MR29]
13. hasp a: b at+ham [MR35] 10. yadaAyix [MR31]
14. haspa: ( at+am [MR36] 11. yadayix [MR35]
15. haspa:"atam [MR37] 12. yadayix [MR37]
'
16. haspa: 'atam [MR42]
17. haspa 'atam [MR44]

-64-
 8. -
ml+XnY 3+a--a+[+!+e_+Lal

l. si.Pr+~+e_+La3+~ [MR3] 1. ml+XanaY

3 +~+!+e_+Lal
2. s:iPr+P+S+h+F [MRlO] [MR3]
3. siPar+P+S+h+F [MR12]
-
4. siPar+im+A+S+h+F [MR12'] [MR9]
5. siPar+im+A+S+ha [MR13.8] 3. ml+XnaY [MRlO]
3 +~+~+£+i
6. siPar+eY 2+A+e_+ha [MR20] 4. ml+XnaY
3 +imA+~+e_+i
7. siPar+eA+S+ha [MR22] [MR12']
8. sPar+eA+S+ha [MR23] 5. ml+XnaY +o:tA+e_+i
3
9. sPareAha [MR3l] [MR13]
10. sfareAha [MR34] 6. -ml+Xn+o:tA+S+i
- [MR15]
[MR35] 7. m+ Xn+o: tA+S+i [MR17]
8. m+Xn+o:taY 2+A+e_+i
[MR20]
9. m+ Xn+o: tay Ae_ [MR22]
10. miXno :tayAe_ [MR26]
11. maXno: tayAe_ [MR30]
12. maXno: tayA [MR31]
13. maXno:tay [MR35]
14. maxno:tay [MR41]
15. maxnotay [MR44]

7. Omissions

The following classes of words are omitted from consideration

in the foregoing morphological statement.
1. Words which are more efficiently constructed
without vowel patterns, including loan words and certain
-65-
common nouns ('"av', '"ax", etc.). Also omitted are
roots with more than three consonants (see p.22).
2. All cases of morphemic alternation, including
morpheme sequence alternation !e.g., nouns which contain
F only in the plural).
3. Certain purely literary and very rare forms of the
verbs of the form CY 2 C: (the so-called 'mediae geminatae').
4. All cases of forms which are not introduced in the
syntactic section. Included here are certain common forms
(e.g., "bayta"), as well as many rare literary forms
(e.g., pronominal verb suffixes, infinitive construct forms,
etc.).
Other occasional omissions and exceptions have been
noted in the text, although not exhaustively or systematic-
ally.

-66-
Footnotes

1. Such considerations are in general not trivial or

'merely esthetic'. It has been recognized of philosophical

systems, and it is, I think, no less true of grammatical

systems, that the motives behind the demand for economy

are in many ways the same as those behind the demand that

there be a system at all. Cf. Goodman (1943).

2. Though not necessarily finite. Thus the resulting gram-

mar will in general contain a recursive specification of

a denumerable set of sentences.

3. Cf. Harris (1951), Appendix to 20.3. A 'grammar of

lists' results from the process of discovery conceived as

the process of determining the extension of the terms

('phoneme', 'morpheme', etc.) which are defined (or, at

the present state, for which procedures of discovery are

given) in linguistic methodology.

4. This is clearly brought out in Harris (1951), Chapter

12, and Hockett (1950).

5. For an elaboration of the point of view sketched here

and its methodological consequences, see my paper "Some

Comments on Simplicity and the Form of Grammars" (unpub-

lished).

6. This formulation explicitly and arbitrarily isolates

morphophonemics for special study. More generally, a com-

-67-
plete grammar in these terms could be described as a single
set of 'transformation statements' generating, from the
most general representation of any sentence, i.e., 'sen-
tence', the specific representations, i.e., all grammatical
phone sequences.

7. See Greenberg (1950) for a discussion of morpheme con-

stituency.

8. Actually the ordering imposed is only partial (see p.

39ff.).

9. See Chomsky (unpub.) for a fuller statement of the

problem.

10. The account of the actual form of the grammar, and the
interpretation and explanation of the notation is impre-
cise and oversimplified, but I think adequate to provide
understandability. It could be given formally, but this is
premature in the absence of a more general study of no-
tations. The notations themselves are not developed in
general form, as they would have to be if presented for
grammar in general, but are specifically adapted to the
needs of this particular grammar.

11. Those morpheme designations which will appear in the

morphology, and which are not given in morphophonemic
spelling, will be underlined to distinguish them clearly
from sequences of morphophoner,les.

12. Common 'compound nouns' of the form N

1 LC~N 1 Lc are often
-68-
treated as a single N , and the ha is prefixed to the first
1
N1 rather than the second, giving forms like "habetsefer",

"haxadar'oxel" instead of "bet hasefer" ('school), "xadar

ha 1 ~xel" ('dining room').

13. In particular, i t follows from this that the definite

article ha is followed by #. This is necessary if, as

seems to be the case, there is a contrast between such forms

as "ha#cdaka" ('the charity') and "hacdaka" ('justification'

If there were no contrast, it would be possible to omit

the juncture (following the orthography) and take the defi-

nite article morphophonemically as haN, although this would

complicate some of the statements of the grammar.

There is one point at which a real internal juncture

might have to be introduced, although this is not done

below. Consonant clusters seem to be all voiced or all

unvoiced, e.g.

hisbir____,. hizbir

hitnagsu~ hitnaksu,

etc., but before the verbal suffixes ti, ta, tern, ten,

consonants seem to preserve voicing, e.g., "la'agti",

"gan~vta". If this is true then a juncture must be intro-

duced before personal suffixes of the verb. Further in-

vestigation is necessary on this point. (I am indebted to

Haim Blanc for information concerning this situation.)

14. This formulation illustrates a case in which a decision

was made to simplify the morphology at the expense of the

syntax. The morphology is simplified if N b£U+ or M+, but

1
-69-
the syntax is simplified if N1 bcM. If this formulation were
accepted, (v) could be dropped from S6, M+ and u+ could

be dropped, and S8 could be reformulated:

S8*. Let M*={Nn+LC(~(Lak+LC))} [n=l,2]

Then Sll* will be:

N {N (M) M t h-+R{a--a; .. etc. }
1= s 1 ' O... e c., i--:e ... etc. '

uo···etc.}

Specification in the morphology of the relevant environ-

ments for a given transformation would then be more com-

plex.

15. I.e., i f there is an LCi in env. [a·Y

~ 2 __ , then J
!~ [2 ~r 4]
throughout. Thus all segments of any one

long component become the same morpheme.

16. Designated 'L' because these are the forms taken by

Lakin certain positions. Lc~ 1 (see MR4).

17. It is apparently disappearing as a separate subset,

partly through analogy (yasan, yosen--see MRl), partly

through disuse in the distinctive forms (as verbs).

18. Many of the members of VlPlb are rare, as is u--0 of

VlP a. Patterns V--0 are traditionally called 'segholates'.

1

19. This gives the purely literary form "nxun~ti", etc.

-70-
20. Only sometimes, see fn. 17, above.

21. This formulation eliminates the possibility of a form

"heCCeC6t". Such forms may arise from an alternation

heCCeC~heGCeC~, the latter occurring before plural (e.g.,

the Talmudic form "hekdesot"). Although I have been unable

to find any such alternations for this form of stems in

Modern Hebrew, they are fairly common with other stems.

If morpheme alternation were considered in the grammar,

then the statement of this alternation would have to follow

2, where the F occurring is the Long Component.

22. Also in env. uc __ in the case of the quadriliteral

3
'smr, giving "'asmura" alongside of "'asmoret". See

p.22.

23. The parenthesized transformation, when applied, leads

to the rare and apparently disappearing second and third

person feminine plural "tiCCoCna", etc., in Verb Future.

In addition, the following transformation sometimes holds:

giving such forms as "tikatavna", "teladna", "hoda'na", by

MR30.5.

24. The forms in which~~ FA are related to the verb

forms, generally, forms like "kabala" being related to the

0+i--:e form (Pi'el), forms like "smira" being related to

-71-
 ~

the 0+a--a form (Kal).

25. Thus "sixva", "kirva" must be derived from sKv, krv,

alternants of sKB, krB, respectively.

26. There seems to be some fluctuation in such forms

even perhaps for one speaker. I cannot at present state
the situation more explicitly.

27. Only sometimes in env. X This is a complex and per-

haps fluctuating situation which requires further investi-
gation for a complete specification, although it could be
given more explicitly than this. Apparently never in M
except in env. __c #.
3

28. Except when R=r•Y , which occurs in this form as

3
"her'a", etc.

29. Exceptions to (6) are "b6hen", "'6hel", "l~xem",

"rexem", and forms from the root hyY (e.g., "yihye").
3
"mixya", "yixye", etc., can be derived from the root
xyY (x.iG).
3

30. As an alternative treatment, the first two lines of

this could be extracted and given in a separate statement
with MR27.2. MR38 could be incorporated into this state-
ment. This would lead to certain changes in the ordering.

31. Exception: "d~ysa". Other exceptions which are not

included in this study for other reasons, are "l§yla",
"habayta", "xaydak". See section 7.

-72-
32. This formulation is tentative. Further research is

necessary to give an authoritative account of the occurrence

of ey and e, but, although the situation may turn out to

be more complicated, it seems to me now that operation and

non-operation of (6) correspond to two dialects. (But for

certain words, e.g., "bet" as in "betsefer", it apparently

holds universally.)

33. It might be necessary to make further mechanical ad-

justments in the output to provide a more adequate phonemic

transcription.

-73-
Bibliography

Chomsky, N. (unpublished). "Some Comments on Simplicity

and the Form of Grammars".

Goodman, N. (1943). "On the Simplicity of Ideas", Journal

of Symbolic Logic 8.4, 107-121.

Greenberg, J. (1950). "The Patterning of Root Morphemes

in Semitic 11 , Word 6.2, 162-181.

Harris, Z.S. (1951). Methods in Structural Linguistics,

University of Chicago Press, Chicago, Illinois.

Hockett, C. ( 1950). "Which Approach to Linguistics is

'Scientific'?", Studies in Linguistics 8, 53-57.

-74-