چكيده لاتين :
1. Introduction
Researchers used to think affixes are meaningless or their meaning is undetermined,
or they obtain meaning by joining other lexical elements (Keshani, 1993; Shari’t,
1372). But today linguists know affixes have semantic parts in the language system.
2. Theoretical Framework
Each lexical semantic representation is composed of two parts: a
Semantic/Grammatical Skeleton and a Semantic/Pragmatic Body. Each skeleton
contains one/more semantic features (show the semantic role of the lexical item) and
one/more arguments. In derivation and compounding the arguments are tied together
via the Co-indexation Principle (Lieber, 2004).
Lieber (2004, 2009) proposes the semantic features [Material], [Dynamic], [IEPS]
(Inferable Eventual Position or State]), [Location], [Bound] and [CI] (Composed of
Individuals) and [Scalar].
[+/- material]: The presence of this feature defines the conceptual category of
substances/things/essences, the notional correspondent of the syntactic category
Noun. The positive value denotes the presence of materiality, characterizing concrete
nouns. Correspondingly, the negative value denotes the absence of materiality; it
defines abstract nouns.
[+/- dynamic]: The presence of this feature signals an eventive/situational meaning,
and by itself signals the conceptual category of situations. The positive value
corresponds to an event or process, the negative value to a state (Lieber, 2004).
The semantic body contains at least two layers. The first one is relatively systematic,
consisting of those universal semantic features being syntactically inactive. This
layer of meaning is relatively stable from speaker to speaker. The second layer of
meaning is purely encyclopedic, consisting of assorted bits of information: color,
precise shape contours, and so on (Lieber, 2009). The skeletons and bodies of author
and bed are shown in (1) a
Journal of Linguistics & Khorasan Dialects Biannual, Vol 8, No. 2 (2016-2017) 2
(1) author Skeleton: [+material, dynamic ([ ], [ ])]
Body: (First layer) < +animate>, < +human>,
(Second layer) {writes for publication, . . . }
(2) bed Skeleton: [+material ([ ])]
Body: (First layer) <-animate>, <+artefact>, <3dimension>,
,
(Second layer) {for sleeping, contains comfortable surface, . . . }
According to the Co-indexation Principle in a configuration in which semantic
skeletons are composed, co-index is the highest non-head argument with the highest
(preferably un-indexed) head argument. Indexing must be consistent with semantic
conditions on the head argument, if any. In the case of a headless structure, co-index
is semantically compatible with arguments (Liber, 2009).
3. Methodology
This paper takes an analytical approach to study the -e Suffix in Farsi. The data
gathered from modern written form Persian and have been analyzed based the
framework presented in Lieber (2004).
4. Results and Discussion
The most productive role of -e is to make concrete nouns, carrying patient meaning,
by attaching transitive verbs. (3) is the -e skeleton, (4) is ferestadan (to send) skeleton
and (5) is the derived item skeleton, ferestade (sendee).
Given the Co-indexation Principle, the highest base argument is co-indexed with the
highest affix argument, provided that the requirement of the affix argument to coindex is regarded. The requirement of the affix argument is non-volitionality; it
means that the affix argument is co-indexed with a non-volitional base argument.
The internal argument of ferestadan is non-volitional; so the two arguments are coindexed. The patient reading of the derived item is the outcome of the co-indexation
of affix argument with the internal argument of ferestadan. It is not the semantic
contribution of -e.
Sometimes the attachment of -e to the verbal base leads to an instrumental reading
(gire (clip), pooshe (file)). (6) is gereftan (to hold) skeleton and (7) is that of gire,
which is the result of joining -e to the present stem of gereftan.
Based on the Co-indexation Principle, the highest arguments of the stem and the
affix are joined provided that the requirement of the affix argument is considered.
(3) -e [+material, dynamic ([nonvolitional ] )]
(4) ferestadan [+dynamic ([ ], [ ])]
(5) ferestade [+material, dynamic ([i-nonvolitional ], [+dynamic ([ ], [i ])])]
-e ferestadan
(6) gereftan [+dynamic ([ ], [ ])]
(7) gire [+material, dynamic ([i-nonvolitional ], [+dynamic ([i ], [ ])])]
-e gereftan
Journal of Linguistics & Khorasan Dialects Biannual, Vol 8, No. 2 (2016-2017) 3
Since the external stem argument in such a context is non-volitional, it gets coindexed with the affix argument and the outcome is a derived item with instrumental
reading. Then the agent/instrumental reading of the derived item is not originated
from the semantic contribution of the affix but the application of the Co-indexation
Principle.
The suffix -e attaches to past stem and sometimes to present stem (xaze) of
intransitive verbs like mordan (to die) with the skeleton in (8) creating a derived
noun with agent reading. (9) is morde (dead) skeleton.
Morde is a person who has done the activity of mordan; but the sole argument of the
predicate has not done the activity willingly and volitionally. Hence this argument is
compatible with the requirement of the affix argument; it is co-indexed with it. The
general interpretation of the derived noun is a non-volitional agent.
The suffix -e is also attached to noun stems and derives a noun with the “similarity
to the stem” reading. (10) is pa (foot) skeleton and (11) is the derived noun paye (leg)
skeleton.
The second argument of pa (its owner) is volitional and inappropriate to be coindexed with the affix argument. Then the affix argument is co-indexed with the first
argument of pa and leads to deriving an item, meaning similar to pa which has the
semantic features [+material] and [+dynamic]. What is the origin of the similarity
reading?
Whenever a language, in the real world, needs a lexical item but she doesn’t have
any derivational means to create such a word either some sort of roundabout process
(e.g., conversion) is employed, or the semantically closest productive affix is put to
use, even if it requires a violation of the Co-indexation Principle in the process
(Lieber, 2004,pp.73-74). Farsi does not have an affix to create the conception of
similarity in words like paye and cheshme (spring); hence one of the most productive
affixes is used. The question is why is not an affix like -i utilized? For us the reason
is the non-volitionality requirement of the affix argument. There is a close
relationship between similarity and non-volitionality.
In sefide (adjective + -e), sefid has a skeleton like (12). (13) is the skeleton of sefide.
The affix argument, regarding its requirement, is co-indexed with the sole argument
of sefid that is non-volitional. Since the affix with [+material] and [+dynamic]
(8) mordan [+dynamic ([ ])]
(9) morde [+material, dynamic ([i-nonvolitional ], [+dynamic ([i ])])]
-e mordan
(10) pa [+material ([ ], [ ])]
(11) paye [+material, dynamic ([i-nonvolitional ], [+material ([i ], [ ])])]
-e mordan
(12) sefid [-dynamic ([ ])]
(13) sefide [+material, dynamic ([i-nonvolitional ], [-dynamic ([i ])])]
-e sefid
Journal of Linguistics & Khorasan Dialects Biannual, Vol 8, No. 2 (2016-2017) 4
features is the derived word head, the outcome of the derivation is a concrete process
noun.
The suffix -e also joins the present stem of transitive verbs and derives nouns bearing
abstract and process reading but lacking agent, patient or instrument interpretation
(xande, larze, gerye). The suffix -e participates in deriving such nouns with a
different skeleton (14). (15) is xandidan (to laugh) skeleton combined with (14) and
derived xande (16).
Since each element has just one argument, the only option is co-indexing these two.
The outcome is an abstract noun bearing process reading; since the head has [-
material] feature and both head and non-head elements have [dynamic].
A noun skeleton like bahār (spring season) (17), combines with the adjective maker
-e (18) and results in deriving an adjective like (19).
The “- affix is not the residue of - Old Persian; but it is the residue of - in
Middle Persian that itself is descendant of - in Old Persian (Khanlari, 1995).
For us the definite maker -e origin is -. Hence, -e has at least two different
origins.
5. Conclusion
The -e has three skeletons: one deriving concrete nouns; but sometimes as a result of
differently applying co-indexation principle and meaning extension it leads to
coinage of nouns with patient, agent, instrument and stem-like readings. One that
derives abstract nouns; the third one derives adjective. The suffix -e has
diachronically two origins: the first one is the -e itself and the other is definite maker
affix changed into -e. This definite maker affix is the source of diminution,
perjoration and endearment.
