Biblical Hebrew Forum

[Ian Hutchesson <mc2499 AT mclink.it>]

Still on the side issue of the English verb system with Rolf in an attempt
to get at a clearer understanding from which to make analogies.

>>                Unmarked    Marked
>>---------------------------------------------
>>  tense         simple      past
>>  complete      simple      continuous  (= incomplete)
>>  at ref. time  simple      perfect (= prior, but tied to rt)
>>---------------------------------------------
>>
>>>(1) Ian has come.
>>
>>The termination though is not supplied by the perfect, but by the
>>non-continuous form of the verb. Consider:
>>
>>  1b) He has been working on his magnum opus.
>>  1c) He has been working on his magnum opus all day.
>>
>>There is no termination involved: we just know from the perfect that what
>>is being described is before now and the continuous (ie not the unmarked
>>form) tells us that it is incomplete.
>
>RF
>By your (1c) you show explicitly that (1b) is contunuous. Does this
>contradict Broman Olsen's model of perfect as perfective? No. 

I may have given examples without adequate explanation. Yes, they are both
continuous in form. They may be perfective, though this has not so much to
do with the verb forms as to each sentence as a whole. I supplied an object
in each case to show that the actions could be completed, but were not so.
While both actions are incomplete, my 1c shows that my 1b is no longer in
progress. But let me give slightly simpler examples

      2a) (Bill's in the shower) he's been working hard.
      2b) (Bill's in the garage) he's been working hard all day.

There is not much hope of these verbs being perfective for they give no
possible closure of the activity. In 2a Bill is specifically not working at
the reference time, though whatever he'd been doing can't be seen as
getting to any conclusion. 2b says that the activity has reached the
reference time.

>Broman Olsen
>would analyze (1b) and (1c) as (C=RT, +imperfective, +perfective). She
>quoted Quirk,R.,Greenbaum,S.,Leech,G., and Svartvik,J., 1985, "A
>Comprehensive grammar of the English language", London: Longman, p 210:

While Quirk and Greenbaum are relatively old time descriptive grammarians,
Geoffrey Leech has attempted a very full systematic componential analysis
of the English language. I also seem to remember another work just by Leech
and Svartvik, a light green book, which was much more progressive than QGLS.

>"/W/hen the perfective and progressive aspects are combined in the same
>verb phrase (e.g.has been working), the features of meaning associated with
>both of them are also combined. Nevertheless, the perfective progressive
>has a semantic range that is not entirely predictable from the meaning of
>its components."
>
>According to Broman Olsen's model the meaning *is* predictable. 

I would in my "humble" way *tend* to agree with this although invariably
combinations will develop a fine patina of significance for their combined
existence.

>
>IH
>>Just out of curiosity how does the tool -- if it is indeed this which
>>supplies your time reference system -- handle
>>
>>  (4)   He said he'd been drinking when the accident happened.
>>
>>given that I'm relating a past in which someone else is relating a past in
>>which something happened prior to that past. (And if Galia is reading I
>>wonder how the system she advocates dealls with it.)
>
>RF
>(4) consiste of three clauses: (1) "He said", (2) "(that) he had been
>drinking", and (3) "when the accident happened".
>If we change the direct speech, we get:
>(5) He said:"I was drinking when the accident happened."

The problem is that this is a snaky sentence which I would translate into
direct speech as:

    3) He said, "I had been drinking when the accident happened."

He's making the accident more coherent for us by giving background
information. However the drinking occurred prior to the accident. This is
why I talked of my shift to the past containing his shift to his past which
at the same time included a prior act to his past. My reference point is
his speech time, while his reference point was that of the accident and the
drinking happened prior to that.

    a) I report to you now that
    b) he said, at an implied point in the past, regarding
    c) the accident which happened, that prior to it
    d) he'd been drinking

a, b and c imply points in time when their actions took place. d does not.
It is connected with the previous point, though it happened sufficiently
close to the accident that its effect was still in operation.

>In contrast with Reichenbach and Comrie, who always take ST (speech time)
>as C (the deictic point) in the simple tenses, Broman Olsen gives evidence
>that C must be pragmatically determined, though the default is ST. In (5) I
>would take ST as C. The main clause "He said" is simple past, and we get
>the formula (RT>C). For the first dependent clause we get a new C connected
>with the saying-incident. Thus ET is the drinking-incident and the accident
>is RT. The expression "was drinking" is past progressive and this means
>that we have an intersection of of ET with RT at the nucleus, i.e. the
>drinking-incident held when the accident happened (the "when-clause" serves
>just as an adverbial). The formula is (RT>C,+imperfective)

I have the feeling that such a descriptive system needs to be mildly
recursive, able to talk of ST1 and we get C1=RT1(=ST2)>ST1 which would be
the equivalent of "He said," C2=RT2>ST2 (implied by the when-clause) "he'd
had an accident", C3>RT2>ST2, "he had been drinking."


I'm now fading so I'll take the opportunity to hit the sack.


Cheers,


Ian