Seminar 3, Week 5

Seminar 3:  Secondary Messages – The Sexy Stuff

Once again we were in tutorial mode, giving an account of the Tense Code in application to Secondary Messages. And thereby completing our account of the Tense Code for English. Next week we will start in on Romance Languages.

Subjunctivitis

Some of you are still suffering from subjunctivitis.  Rather than lay out the full story ourselves, we refer you to a 1988 paper, Indicative and Subjunctive, Analysis 48:3, 113-122, which says (almost) all of what we have to say on the matter.  Those seeking further insight may also care to peruse a 1994 offering, Against the Indicative, Australasian Journal of Philosophy 72:1, 17-26.  Part I of this article contains a few barbed comments about the grammarians, but also lays out the case against philosophers’ deployment of the ‘subjunctive’ in their taxonomy of ‘if‘-sentences.  You might also read this latter piece for revision of Seminar 1: it urges the case against the doctrine that English has a future tense with ‘will‘ as its marker. 

No doubt subjunctivitis will surface again when Agent Kelly speaks in Week 7, so the above are offered merely as prophylactics.

Secondary Messages

We distinguished three categories of Secondary Message, Proper Messages (which Brunellus suggested we might rechristen PROPENSITY Messages, Practical Messages (Which Brunellus suggested we might rechristen HISTORICAL Messages), and Projective Messages (where we are entirely happy with the label).

These three are distinguished grammatically by the role of tense, which although invariably marking a piece of purely temporal intelligence, marks different pieces of temporal intelligence across the categories.

Let us remind you that in Primary messages, tense is identical to time-about. So primary messages are PROPOSITIONS, claims of fact about their tenses, and are true or false depending on whether or not the predication condition is satisfied at the appropriate time. ‘Time-about’ begins as an intuitive notion, the time that a message is intuitively ‘about’. But we give it a precise technical formulation. It is the time p of the satisfaction of the predication condition.  So for Primary messages, we write  t = p

And now the headlines of our tripartite classification of secondary messages.

In Proper messages, tense is again identical to time-about. So proper messages too are propositions, claims of fact about their tenses.  Again we write t = p

Examples for definiteness.  Each variant of

Her Majesty {will (often)/may (occasionally)/can (sometimes)} be found sipping a sherry in Balliol Buttery

has an interpretation deponing Her Majesty’s present habit, tendency, propensity (hence Brunellus’ suggestion), call it what you will, concerning sherry in Balliol.  The aboriginal forms encode for presentness, and the message as a whole is true or false depending on the facts at p = 0.

And each variant of

Her Majesty {would (often)/might (occasionally)/could (sometimes)} be found sipping a sherry in Balliol Buttery

has an interpretation deponing Her Majesty’s past habit, tendency or propensity, her propensity at some p < 0.  The transported forms encode for pastness, and the message is again true or false depending on how the facts lie at p.

N.B.  Proper messages deploy only the first two elements of our tense code.  There is no past-past version of these messages, and so no forms deploying transportation + phase-modification in that order.  Nor would we expect any, for propensities are states: they hold at a time, either present or past.  And propensities which are alleged to hold at a point before some salient past moment are just past propensites.

Now for Practicals and Projectives.  The first point to note is that both Practical and Projective messages encode not propositions but JUDGEMENTS, tensed verdicts.  Judgements are not meant to hold come what may, but in the normal course of events.  English offers two kinds of judgement.  Judgements concerning matters of historical fact, past or present.  These are the Practicals   And judgements concerning situations that can only be imagined. These are the Projectives.  

Tense plays a different role in the two categories.

In Practical messages, tense marks the time of the STANDPOINT  s from which the judgement is ventured. Time-about, the time of predication, is encoded quite separately, by presence or absence of phase. Here tense and time of satisfaction of the predication condition may coincide, but they are separate matters, encoded independently.  So t = s, present or past, according as aboriginal or transported forms of the modals are used.  When  p < 0, the sentence is phase-modified.  When p = o, it is not.

Now this idea of a standpoint s is a subtle one, and it may be easier to grasp once you have something to compare it with, so we suggest looking at Projectives first.

But let me sketch an account (you will find the detail in the series of posts beginning with Most Australians).  Any  judgement will be supported by reasoning, reasoning which justifies the verdict.  And that reasoning may rely upon an up-to-the-minute assessment of the world and its ways.  s is then the point of speech, and we write s = 0.   But the reasoning may rely only upon past experience, and information gained thereby.  Whereupon there is the option of signalling that fact, explicitly disowning reliance on the latest news, by selecting a transported form of the modal, locating  s  in the past.  We write s < 0

Examples, to help fix the idea.  Each variant of

Grannie {will/won’t/may/mayn’t/must/can’t/needn’t/daren’t} be in the attic

has a Practical interpretation concerning Grannie’s present whereabouts, viewed from a present standpoint.  Here ts = 0 (encoded by the aboriginal form of the modal) and p = 0 (encoded by absence of phase). 

Each variant of

Grannie {will/won’t/may/mayn’t/must/can’t/needn’t/daren’t}  have been in the attic

has a Practical interpretation concerning Grannie’s past whereabouts, viewed from a present standpoint.  Here ts = 0 (encoded by the aboriginal form of the modal) and p < 0 (encoded by presence of phase). 

Likewise, each variant of

Grannie {wouldn’t/might/might not/could/couldn’t} be in the attic

has a Practical interpretation concerning Grannie’s present whereabouts, viewed from a past standpoint.  And thus disowning any reasoning based narrowly on present circumstances. Here ts < 0 (encoded by the transported form of the modal) and p = 0 (encoded by absence of phase). 

And finally, each variant of

Grannie {would/wouldn’t/might/might not/couldn’t} have been in the attic

has a Practical interpretation concerning Grannie’s past whereabouts, viewed from a past standpoint.  Here ts < 0 (encoded by the transported form of the modal) and p < 0 (encoded by presence of phase). 

Note, before we leave Practicals, that just as with Proper messages, they use only a stripped-down version of the tense code.  There are no past-past practical messages.  Indeed there could not be, for phase is given over to identifying p, the point of predication.  Leaving no resources available for encoding a past-past standpoint. 

Now for the Sexy Stuff.  It is with Projectives that the story gets exciting.  For a start, they use the full range of the tense code, giving present, past and past-past varieties. And understanding them requires an ability to distinguish what is fact and what is fantasy. The headlines:

In Projective messages, tense marks the CHANGE-OVER POINT, c the point at which the deliverances of history cease, and the imagination takes over. In projective messages tense is always earlier than time-about.  And all possible combinations consistent with that defining characteristic (t < p)are available.

We may have present-tensed messages concerning the future satisfaction of a predication condition (t = c = 0; p > 0), past-tensed messages concerning the future satisfaction of that condition (t = c < 0; p > 0), and past-past-tensed messages concerning the future satisfaction of the same condition(t = c << 0; p > 0).  The natural interpretations of these remarks, made on September 1 1940, each concerning a future German victory in the aftermath of an invasion, are cases in point:

If Hitler invades, Germany {will/may/can} win the war
If Hitler invaded, Germany {would/might/could} win the war
If Hitler had invaded, Germany {would/might/could} have won the war

We can also have (t = c < 0; p = 0), past-tensed messages concerning present satisfaction of  a predication condition:

If your father was alive today, he would be turning in his grave

And also (t = c <, 0; p = 0), past-past-tensed messages concerning the present:

If your father had been alive today, he would have been turning in his grave.

And to complete the pattern we can have (t = c << 0; p < 0), past-past-tensed messages concerning the past satisfaction of a predication condition.  As with a remark made now about the last war:

If Hitler had invaded, Germany {would/might/could} have won the war

That’s enough from us.  The rest of the content of Seminar 3 has already been laid out by one who came before us.  It is in Dudman’s 1994 gem: On Conditionals. All laid out pretty, with diagrams.  He doesn’t here use the terminology of a change-over point c, preferring instead the date D of the fantasy’s inception.  But it is exactly the same notion under another name.

And that’s it for Theory.  Next week we make a start on the real business of the project, as we begin to investigate whether, and how far, the theoretical grid we have laid out for English maps on to other languages.  First, Romance.  The introduction to Seminar 4, once we have time to write it, will specify some tasks which may be fruitfully addressed in advance.