IV-1 The Arabic verb composition The Arabic verb is a dense morphological composite. Its root carries the lexical/semantic meaning. Tense, aspect, and agreement denotation of its affixes and/or vocalic melody is disputable. In traditional Arabic grammar, the verb inherently denotes a meaning in relation to ibn hiām, n.d.:14). It) “الفعل ما دل على معنى في نفسه مقارن بأحد األزمنة الثالثة” one of the three tenses has three forms: الماضي [mādī] past: what has elapsed, األمر [amr] imperative: what is going to be and has not happened, and المضارع [mudāri] present: what is being and not having ceased )ibn hiām, n.d.:20; sībawayhi, 796 cited in Bahloul, 2008:40). The three forms differ morphologically, while the future, which is not included as a separate form, is morphologically identical to the present, and is often lexically marked by سوف [sawfa] will or its alternative prefix [sa-]. In this classification, the temporality of the imperative derives from the potential 58 occurrence of the event in future time. This view is adopted in modern Arabic grammar books and by some linguists (Bahloul 2008:40). Others claim that the Arabic verb is not temporal, but aspectual. It has two forms, the first denotes “a finished act, one that is done and completed in relation to other acts (the Perfect); the other an unfinished act, one that is just commencing or in progress (the Imperfect)” (Wright 1996:52). Based on the notion of abstract vs. relative tense, Comrie

Based on the notion of abstract vs. relative tense, Comrie (1976) argues that Arabic verb forms have a combined Aspect/Tense meaning; where tense is relative to some process other than current time. Hence, the “perfect refers to past time (or completed action) and the imperfect to present or future time (an incomplete action)” (Bateson, 2003:23). This dual tense-aspect characterization of Arabic verbs is also forwarded by Fassi Fehri (1993) based on a view that verbal inflections have a combined grammaticalization of aspectual and tense oppositions. From a different perspective, Aspect/Tense are considered abstract; i.e. not realized morphologically and affixes are only agreement morphemes. This view derives from the observation that the negative verb ليس [laysa] is not and the aspectual مازال [mā-zāla] is still denote present, but inflect with perfective suffixes. Similarly, the vocalic melody of the verb does not carry tense or aspect, since these negative and aspectual particles resemble perfective verb vocalic melody (Aoun, et al., 2010:22-3). Aspect/Tense of the imperfective are also abstract and their denotation varies widely as will be shown in IV-2 & IV-3 (Aoun et al., 2010:24-6). Additionally, Arabic dialects have independent aspectual markers that attach to the imperfective, 59 like [b-], ammāl], [baa], the participle, the copula, and adverbial complements (Mitchell & alHassan, 1994).

...................

1 Aspectual Approaches: An Introduction As verbs denote events that take place in time, can verbs be differentiated according to how the events they denote take place in time? THE PROPOSAL: The ontological types of events are determined by their temporal contours. That is, aspectual notions are critical to structuring lexical semantic representation. THE BIG QUESTIONS: —

How are verbs classified with respect to the internal temporal structure of the events they denote? —

 What evidence is there for determining a verb’s class membership? —

What is the best characterization of the relevant aspectual notions?

Whereas tense is deictic, placing a situation in time, aspect informs about the contour or the quality of the event or the state as seen by the speaker. Aspectual meaning is a composite of the information provided by the so-called situation-type aspect and the one provided by the so-called viewpoint aspect (Smith 1991). Situation type aspect is determined by the inherent properties of situations: they can be durative, instantaneous, telic or atelic. This information is actually given by the lexical verb and its arguments, as well as by other complements present in the sentence.

• Most means/manner verbs describe events that take time but have no inherent temporal endpoint built into them; these events could go on indefinitely — at least, if real world limitations or conventions weren’t a consideration. Such events are known as ACTIVITIES. •

Most result verbs also describe events that take time, but these events have an inherent temporal endpoint: the time when the result state comes about. The duration of the event up to the endpoint is taken up by the process that leads up to this result state, though it is left unspecified by the verb; it is this process that a means/manner verb names. Such events are known as ACCOMPLISHMENTS.

• Some means/manner verbs, such as hit, describe events that are punctual — they take no more than a moment in time (Engelberg 2000)— but as with the events described by durative means/manner verbs there is no result state that follows. Such verbs can name a single hit or a series of hits; on the latter interpretation, there is again, no inherent endpoint specified to the hitting. Such events are known as SEMELFACTIVES.

• Some result verbs, such as break, describe events that are punctual, but the moment described involves the transition to the result state, and there is often a sense that an accompanying process is presupposed. Such events are known as ACHIEVEMENTS.

3 Lexical Aspectual Classes of Verbs ASPECTUAL NOTIONS INTRODUCED SO FAR: — duration/punctuality — telicity, i.e., having a culmination or endpoint ONE OF THE MOST BASIC ASPECTUAL NOTIONS: Stativity or its inverse, dynamicity — events that involve change: activities, accomplishments, achievements, semelfactives run, walk, sleep, sweep, reach, win, break, knock, hit — events that do not involve change: states hate, know, believe, be red THE RESULT: FIVE major aspectual classes.

THREE DESCRIPTIVE PROPOSALS CONCERNING HOW BEST TO CHARACTERIZE STATES • Statives, unlike nonstatives, can be judged true at a moment in time (Dowty 1979) INADEQUATE: so-called “interval statives”: spatial configuration verbs such as sit, stand, lie. (4) a. My keys are sitting on the shelf. b. The statue of George Washington stands in the Common. • Stative verbs involve no change, nonstative verbs do (Dowty 1979, Kearns 1991, McClure 1994); i.e., activities are iterated changes. (5) The general observation is that states have no essential changes or transitions, from which it follows that they are continuous and are not essentially bounded. (Kearns 1991:116) • Nonstates, but not states, require energy input to maintain event (Comrie 1976:48).

3.2 Systems of aspectual classification There are various aspectual classification systems, though they make essentially the same distinctions, collapsing some or subdividing others (Bach 1981, 1986, Dowty 1979, Mourelatos 1978, Vendler 1957). SOME MAJOR DIMENSIONS OF VARIATION:

 • Semelfactives are often neglected and subsumed under achievements, as they also are punctual. THE CONSEQUENCE: Many systems of aspectual classification recognize only four aspectual classes, the so-called Vendler classes: activity, accomplishment, achievement, state.

 • Some researchers privilege the distinction between verbs denoting events with an inherent endpoint (telic) and those without (atelic) over other aspectual distinctions. • Some researchers argue that the accomplishment/achievement distinction is pragmatic and not linguistic, thus collapsing these two classes (e.g., Verkuyl 1993). (7) a. Taylor ate a peach. b. Taylor ate a grape

 • Some researchers recognize two subclasses of achievements: achievements with an associated process and purely “lucky” achievements

------

As it is not a natural language, MSA morphology has been described as an extremely regular system [5], susceptible of being represented by means of precise formal devices. As Kaye describes it, MSA presents an “almost (too perfect) algebraic-looking grammar” [2:665]. Broadly speaking, stems—word-forms without the affixal material [6]—are generally built by the organized combination of two types of morphemes—what we call the root and the pattern. The MSA lexicon contains between 4000 to 5000 different roots [7,8], and verbal morphology exhibits 24 different patterns, of which 16 are really common. Semantically related words tend to share the same root morpheme. Thus, the root turned out to be the basic component of Arabic lexicography, to the extent that dictionaries are organized by roots [9]. At a more superficial level, the inflectional system applies several operations to turn stems into specific verbal wordforms. This stage is considerably complicated by the interaction of phonological and orthographic alterations. All these phenomena hinder the process of formalizing the sytem, thus making it an extremely interesting and challenging task.

Templatic morphology is known in the field of Arabic linguistics as root-and-pattern morphology. It takes its name from the Arabic morphemes which have a non-concatenative shape: the root and the pattern. This theoretical description attempts to describe how Arabic stems are built—root-and-pattern morphology states that stems are composed by these two elements. A root is a decomposable morpheme that provides the basic meaning of a word, and generally consists of 3 or 4 ordered consonants in non-linear position within the word [10,11,12,13,1]. The pattern is a syllabic structure which contains vowels, and sometimes consonants, in which the consonants of the root are inserted and occupy specified places [14,15]. Thus, by the interdigitation of a root and a pattern stems are created [16,17,18,10,15]. Some authors have proposed to separate the vowels from the template and to consider it a separate morpheme. This morpheme is commonly known as vocalism [19,20,21,22].

MSA exhibits 24 different verbal patterns. Some of them belong in fact to Classical Arabic and are rarely used. Traditionally they are classified in patterns from 3-consonant roots and patterns from 4-consonant roots. The different patterns add extensions to the basic meaning expressed by the root, i.e., they are of derivational nature. Below, we include the list of patterns using the root فعل fçl ‘doing’. This root is traditionally used in Arabic to refer to grammatical forms. Patterns are shown using the lemma of the verb, which corresponds to the third person masculine singular of the perfective active inflection [4,10,18,23,24]. Following the Arabic western linguistic tradition, we use Roman numerals to refer to the different patterns. Patterns I include two vowels in their specification: one corresponds to the thematic vowel of the perfective and the other one to the thematic vowel of the imperfective—both correspond to the second vowel position of the stem. Some verbs share the same lemma form, but they are considered different since they present different forms in their conjugation. 4- consonant roots are distinguished from 3-consonant roots by the addition of a ‘Q’ to the Roman numeral.

The fact that a small number of syllabic structures is allowed by Arabic phonotactics has interesting implications: as the formation of words belonging to the same morphological class is the product of a quasi mathematical combination of similar morphemic material, the resulting syllabic structure will tend to follow the same patterns. Thus, it seems possible to propose a precise formalism which predicts the syllabic structures for Arabic lexical items.

Verb framing

On the other hand, all Romance languages are verb-framed. Spanish, for example, makes heavy use of verbs of motion like entrar, salir, subir, bajar ("go in", "go out", "go up", "go down"), which directly encode motion path, and may leave out the manner of motion or express it in a complement of manner (typically a participle): entró corriendo "he ran in'", literally "he entered running"; salió flotando "it floated out", literally "it exited floating".

Verb framing is also used in some non-Romance languages such as Turkish, Hebrew and Arabic. In Arabic, for example, dakhala rākiḍan means "he entered running", with dakhala meaning "to enter" and rakaḍa meaning "to run".

Since at least the early days of Generative Semantics, one of the primary concerns of lexical semanticists has been the question of possible and impossible verb meanings.

According to Ozcaliskan & Slobin (2000), Arabic is a verb­framed language that tends to conflate motion and path in the verb root. For example, the root verb ɣadarә درñı means to go out. However, English is a satellite­language that conflates motion and manner in the verb root such as the verb scramble while path is mapped onto a satellite (e.g. out, in, up, etc.). All the five English verbs above and their Arabic counterparts are verbs that conflate motion and manner

 that opposes satellite-framed languages, in which “the direction of motion must be expressed by a particle or prefix and not the verb itself,” to verb-framed languages, in which “the direction of motion is encoded in the verb” and “manner of motion must be expressed by another component, i.e. an adverb or a gerundive.”

Self Motion

In order to accumulate sentences that evoke events of ‘Self­Motion’ in English and Arabic, I decided to search for five English verbs as research terms. The English verbs are: walk, run, fly, climb, and crawl. The verbs are among over 140 lexical units (LUs) that evoke the ‘Self­Motion’ frame in English according to FrameNet . I chose these five verbs because 3 of their frequent use as well as their tendency to have monomorphemic Arabic counterparts. Their equivalent Arabic verbs are: مشى maʃa [=walk], جرى ʒara [=run], طار tˤara [=fly], تسلق tʰasaĺqa [=climb], and ŋąز zaħәfa [=crawl]. I have searched and translated the Arabic data myself and also consulted three different online Arabic­English dictionaries .

4.1. Self­Motion Frame In this section, I present a description of the ‘Self­Motion’ frame. This frame is based on the English ‘Self­Motion’ frame as represented in FrameNet. The frame describes the movement of an entity that moves independently without direct external influence. The entity starts the movement from a source point, goes through a trajectory visible or invisible and ends up in at a goal location. The movement takes place within a bounded area and has a specified direction.

TENSE refers to the absolute location of an event or action in time, either the present or the past. It is marked by an inflection of the verb:  David walks to school (present tense)  David walked to school (past tense)  Reference to other times -- the future, for instance -- can be made in a number of ways, by using the modal auxiliary will, or the semi-auxiliary be going to:  David will walk to school tomorrow  David is going to walk to school tomorrow.  Since the expression of future time does not involve any inflecton of the verb, we do not refer to a "future tense". Strictly speaking, there are only two tenses in English: present and past.  ASPECT refers to how an event or action is to be viewed with respect to time, rather than to its actual location in time. We can illustrate this using the following examples:  [1] David fell in love on his eighteenth birthday   [2] David has fallen in love   [3] David is falling in love  In [1], the verb fell tells us that David fell in love in the past, and specifically on his eighteenth birthday. This is a simple past tense verb.  In [2] also, the action took place in the past, but it is implied that it took place quite recently. Furthermore, it is implied that is still relevant at the time of speaking -- David has fallen in love, and that's why he's behaving strangely. It is worth noting that we cannot say *David has fallen in love on his eighteenth birthday. The auxiliary has here encodes what is known as PERFECTIVE ASPECT, and the auxiliary itself is known as the PERFECTIVE AUXILIARY.  In [3], the action of falling in love is still in progress -- David is falling in love at the time of speaking. For this reason, we call it PROGRESSIVE ASPECT, and the auxiliary is called the PROGRESSIVE AUXILIARY.