3.5 Behaviour of the verb
4.1.2 Properties of phasal verbs
This section is concerned with the different ways in which phasal verbs are treated in the
literature, where they can be accounted for as auxiliaries, complex predicates or raising verbs.
4.1.2.1 Phasal verbs as auxiliaries
In his account of auxiliary-hood in Egyptian Colloquial Arabic (ECA), Eisele (1992) treats
phasal verbs as members of the lexical category of auxiliaries. This is based on the notion
that they form compound-verb phrases where they represent the highest verb which carries the
deictic time reference, and in that they resemble k¯an, which is an auxiliary in ECA. This is
specifically true of the phasal verb yibPa ‘to become’, for which Eisele (1992) shows examples
of embedding a perfective verb, a bare imperfective, active participle and a predicate nominal,
this shows that it can indicate a variety of temporal contrasts through embedding most types
of predicates (Eisele, 1992). Therefore, based on Eisele (1992)’s account, phasal verbs represent
the head of a complex verb construction in which they carry the decitic time reference. The
category of auxiliary-hood is also described as a fuzzy category with a continuum of features,
where some members are prototypical auxiliaries, such as k¯an, while other members are more
peripheral, such as modals and phasal verbs (Eisele, 1992).
The verbs investigated in Eisele (1992)’s study as phasal verbs (refered to as aspectualizers)
were: fidil ‘remain’ , PaQad ‘continue’, baPa ‘become’, Q¯ad ‘go back’1, Pidir ‘can’, Parrab ‘be
nearly’, Qirif ‘know how’, lièiP ‘catch’, Pibtada ‘begin’2, è¯awil ‘try’, battal ‘stop’. The verb baPa
‘become’ is also included but it is considered by Eisele (1992) to be a real auxiliary, just as k¯an
‘be’.
However, some of these verbs were not used in the current study, as they do not really represent
phasal verbs either in their meanings or in syntactic behaviour, such as è¯awil ‘try’ and Qirif 1This verb is not used in ECA as a phasal verb as far as I can judge, it is used in MSA. Therefore it was not
included in the current study.
2This phasal verb is investigated in the current study as badaP, where both are different templates for the
‘know’.
There are a number of features that Eisele (1992) uses to describe the behaviour of phasal verbs
as belonging to the category of auxiliaries, these are listed below in order of importance for
distinguishing verbs as belonging to the lexical class of auxiliary verbs:
• They do not subcategorize for a complementiser, and this is listed as the most important
feature of an auxiliary verb. However, badaP ‘begin’, è¯awil ‘try’ and battal ‘stop’ are listed
as having this feature optional, where they might have a complementiser following them
(Eisele, 1992).
• Subject coreferentiality between the phasal verb and the following lexical verb is obligatory.
This feature is shared among all of the phasal verbs, and it is mentioned as the second
most important feature of being an auxiliary according to Eisele (1992).
• The lexical verb following the auxiliary may not have an independent temporal reference
(deictic time reference in Eisele (1992)’s terms). All of the phasal verbs mentioned in his
study share this property of lacking a discrete deictic time reference for its complements.
This description shows that although phasal verbs share some behaviours with the auxiliary k¯an,
they do not form a coherent set as they show different behaviours as is shown above. These
behaviours will be further investigated in the data section of this chapter, to show if they still
hold for the data set used in the current study, and this will then be used to investigate how
these verbs should be analyzed syntactically.
This is close to the description of these verbs in Michaelis (1998) as ‘semi-auxiliaries’, where it
is stated that: “In English, the inceptive aspect is expressed by means of the ’semi-auxiliaries’
begin and start. These verbs qualify as semi-auxiliaries (SAs) because they partake of semantic
properties of true auxiliaries while also exhibiting certain behavioral properties of main verb.
SAs resemble main verbs in the following respects: they do not undergo subject-auxiliary inver-
in each of these cases. SAs share with auxiliaries the following semantic property: they are
subject-raising predicates. That is, the particular SA does not impose any thematic role spec-
ification upon the subject argument, this specification being supplied by the downstairs verb”
(Michaelis, 1998, p.85). This study, however, is based on examples from English, and therefore
the criteria might not apply to the current set of phasal verbs.
4.1.2.2 Phasal verbs as complex predicates
According to (Eisele, 1992), aspectualizers form compound-verb phrases where they represent
the highest verb which carries the deictic time reference, and in that they resemble k¯an. This is
true for verbs such as yibPa where: “yibPa (imperfect of baPa ‘be, become’) is used similarly to
k¯an, in that it can embed most types of predicates to indicate temporal contrasts” (Eisele, 1992,
p.154). Here it is clear that baPa is treated by Eisele (1992) as another form of k¯an which is
a true auxiliary, rather than as a phasal verb. baPa is investigated in more details in section 4.3.1.
Maas (2009) has a similar account, where these verbs are treated as parts of a complex predicate.
They are called ‘coverbs’ and are treated similarly to k¯an which also functions as a temporal
marker for the whole construction.
4.1.2.3 Phasal verbs as raising verbs
In Michaelis (1998)’s account of these verbs as semi-auxiliaries (SAs), the semantic property
that these verbs were assumed to share with auxiliaries is that “they are subject-raising pred-
icates. That is, the particular SA does not impose any thematic role specification upon the
subject argument, this specification being supplied by the downstairs verb” (Michaelis, 1998,
p.85).
eat meat, where Ram is clearly the subject of began. It is generally assumed that, in cases like
(i), the subject of the complement has been raised to be matrix subject [see Newmeyer (1975),
Langacker (1995) for discussion of this issue]” (Noonan, 2007, p.141).
Investigation of phasal verbs as raising predicates is tackled in more details in section 4.4.2 with
respect to the data of the current study.