The History of Branching Time

Peter hrstrm
Department of Communication
Aalborg University
Denmark

How should time be represented graphically?

Circular time

Linear time

Branching time
Now
 How should time be represented graphically?

Branching time
Now

Branching time in literature:

Borges: The Garden of Forking Paths

Hans Christian Andersen: The Story of a Mother
.. probably many others .

Hans Christian Andersen: The Story of a Mother

Future 1 Future 2

She looked into the well, and "One life that you saw
it made her glad to see how was your child's fate,
one life became a blessing to your own child's future."
the world, for it was so kind And she bowed her
and happy. Then she saw the head, as Death took her
other life, which held only child to the unknown
sorrow, poverty, fear, and land.
woe.
 Who invented the idea of branching time?

Henri Bergson A.N. Prior Saul Kripke

(1859-1941) (1914-69) (born 1940)

Who invented the idea of branching time?

. an illustration of the process

of deliberation MO, a series of
conscious states. At the state O
two directions, OX and OY,
equally open. .
Henri Bergson: Time and Free
Will (1910)

This is deceptive: Time as space

Henri Bergson
(1859-1941) We can analyse a thing but not a process; we
can break up extensity, but not duration
(p.219)
 Who invented the idea of branching time?
And I think it important that people
who care for rigorism and formalism
should not leave the basic flux and flow
of things in the hands of existentialists
and Bergsonians and others who love
darkness rather than light, but we
should enter this realm of life and time,
not to destroy it, but to master it with
our techniques.
(Prior, undated note)

Arthur Norman Prior

(1914-69)

Who invented the idea of branching time?

Saul Kripke, Born Nov. 13, 1940.

Prof., Princeton, 1977-1998; distinguished professor of
philosophy at The Graduate School and University
Center of the City University of New York
 Prior, A.N.: Time and Modality, Oxford 1957

Basic axioms in modal logic + PC:

L(p q) (Lp Lq)
Lp p
Mp ~L~p
Rule: If p is provable then Lp is provable
Temporal interpretation of modal logic:
p is an arbitrary proposition the truth-value of which
depends on a descrete and linear time.
Mp means: p is or will be the case
Prior uses 1 for true and 3 for false.
p 1 3 1 3 1 3 3 1 1 3 3 only 3s
after this

Prior, A.N.: Time and Modality, Oxford 1957

S4-axioms added to PC:

L(p q) (Lp Lq)
Lp p
Lp LLp
Rule: If p is provable then Lp is provable
Temporal interpretation of modal logic:
p is an arbitrary proposition the truth-value of which
depends on a descrete and linear time.
Mp means: p is or will be the case
Prior uses 1 for true and 3 for false.
p 1 3 1 3 1 3 3 1 1 3 3 only 3s
after this
 Prior, A.N.: Time and Modality, Oxford 1957

p is an arbitrary proposition true of which

depends on a descrete and linear time.
Mp means: p is or will be the case
Prior uses 1 for true and 3 for false.

p 1 3 1 3 1 3 3 1 1 3 3 only 3s
after this
Mp 1 1 1 1 1 1 1 1 1 3 3 only 3s
after this
Prior suggests that this use of M corresponds to S4.

Saul Kripkes letter to Prior: Sept. 3, 1958

p is an arbitrary proposition true of which

depends on a descrete and linear time.
Mp means: p is or will be the case
Prior uses 1 for true and 3 for false.

p 1 3 1 3 1 3 3 1 1 3 3 only 3s
after this
Mp 1 1 1 1 1 1 1 1 1 3 3 only 3s
after this
Prior suggests that this use of M corresponds to S4.
Saul Kripke rejects this. He shows that LMp LM~p is valid
in Priors system, but not provable in S4.
 Saul Kripkes letter to Prior: Sept. 3, 1958

p is an arbitrary proposition true of which

depends on a descrete and linear time.
Mp means: p is or will be the case
Prior uses 1 for true and 3 for false.
Which temporal logic corresponds to S4?

Now in an indetermined system, we perhaps should not

regard time as a linear series, as you have done. Given the
present moment, there are several possibilities for what the
next moment may be like -- and for each possible next
moment, there are several possibilities for the next moment
after that. Thus the situation takes the form, not of a linear
sequence, but of a tree.

Saul Kripkes letter to Prior: Sept. 3, 1958

p is an arbitrary proposition true of which

depends on a descrete and linear time.
Mp means: p is or will be the case
Prior uses 1 for true and 3 for false.
0
Rank 1

Which
temporal 1 2 3
Rank 2

logic Rank 3

corresponds 4 5 6 7 8 9 10 11 12

etc etc etc

to S4?
etc etc

etc
 Saul Kripkes letter to Prior: Sept. 3, 1958

p is an arbitrary proposition true of which

depends on a descrete and linear time.
Mp means: p is or will be the case
Prior uses 1 for true and 3 for false.
0
Rank 1

a proposition
is considered 1 2 3
Rank 2

necessary if Rank 3

and only if it is 4 5 6 7 8 9 10 11 12

etc etc etc

and definitely
always will be
etc etc
the case.
etc

Saul Kripkes letter to Prior Oct. 13, 1958

Do you think a tensed logic is needed for
scientific discourse? I should think that, for
scientific discourse a tenseless logic may be
preferable.
0
Rank 1

1 2 3
Rank 2

Rank 3
4 5 6 7 8 9 10 11 12

etc etc etc

etc etc

etc
Priors letter to Kripke Oct. 27, 1958
I do not see how indeterminism can be
expressed in a tenseless language at all. For
indeterminism asserts a certain difference
between the future and the past
0
Rank 1

1 2 3
Rank 2

Rank 3
4 5 6 7 8 9 10 11 12

etc etc etc

etc etc

etc

Further development of branching time.

Kripkes possible future corresponds to
MF(x)p. - P(x)p seems easy to represent.
But how can F(x)p, MF(x)p and NF(x)p be
understood in this framework?
0
Rank 1

1 2 3
Rank 2

Rank 3
4 5 6 7 8 9 10 11 12

etc etc etc

etc etc

etc
 Further development of branching time.
Could there be alternatives pasts in a
branching time system?

This would mean that it should be accepted that although

the proposition q was true yesterday, it may not be true
today that it q was the case yesterday. According to Prior
this would not be acceptable.

The asymmetry between past and future

This belief of mine... is bound up with a

belief in real freedom. One of the big
differences between the past and the future
is that once something has become past, it
is, as it were, out of our reach - once a
thing has happened, nothing we can do can
make it not to have happened. But the
future is to some extent, even though it is
only to a very small extent, something we
can make for ourselves.... if something is
the work of a free agent, then it wasnt
going to be the case until that agent
decided that it was.
[undated note]
 Further development of branching time.
Kripkes possible future corresponds to
MF(x)p. - P(x)p seems easy to represent.
But how can F(x)p, MF(x)p and NF(x)p be
understood in this framework?
Ockhams (the medieval) view:
At least one chronicle
is selected i.e. it
represents what
freely will be chosen.
Now In this way MF(x)p,
F(x)p and NF(x)p
will be different.
Problem: How can we account for the difference between two
different kinds of branches?

Further development of branching time.

Kripkes possible future corresponds to
MF(x)p. - P(x)p seems easy to represent.
But how can F(x)p, MF(x)p and NF(x)p be
understood in this framework?
Priors Ockhamistic solution:
True-values have to
be calculated relative
to moments and
c chronicles:
t Ock(p,t,c)
In this way MF(x)p,
F(x)p and NF(x)p
will be different.
 Further development of branching time.
Kripkes possible future corresponds to
MF(x)p. - P(x)p seems easy to represent.
But how can F(x)p, MF(x)p and NF(x)p be
understood in this framework?
The Peirce solution: No future or
counterfactual branches
are selected, but F(x)p
and is simply
understood as NF(x)p.
P(x)q P(x)~q holds.
Now
F(x)q F(x)~q does not
hold.
Problem: How can we account for the fact that F(x)p F(x)~p
is not a theorem is the Peirce system?

The two languages: Priors view

I believe that what we see as a

progress of events is a progress of
events, a coming to pass of one
thing after another, and not just a
timeless tapestry with everything
stuck there for good and all... [Prior,
Undated]
 J.M.E. McTaggart (1866-1925): Two languages about
time

Famous paper: The Unreality of Time [1908]

The A-language: past, present, future

The B-language: before, after, simultaneously with

The semantics (the B-language):

Time is a set of moments with an ordering relation, <,
corresponding to before/after and

Any well formed proposition has a truth-value at a moment

of time.

Definitions of the tenses:

T(t,Fp) def t1 : t<t1 T(t1,p)
T(t,Pp) def t1: t1<t T(t1,p)

Priors Ockhamistic system

Ock(p,t,c) : the truth-value of p at t for the chronicle c
through t
 Two languages about time: The relations

Which of the two languages (if any) should

be considered to be the basic (fundamental)
language in temporal discourse?

Priors 4 grades:
1. The B-notions are more fundamental that the A-notions.
Therefore, in principle the A-notions have to be defined in
terms of the B-notions.
2. The B-notions are just as fundamental that the A-notions.
The two sets of notions have to be treated on a par.
3. The A-notions are more fundamental that the B-notions.
There is also a primitive and fundamental notion of
(temporal) possibility. In principle the B-notions have to be
defined in terms of the A-notions and the primitive notion of
temporal possibility.
4. The A-notions are more fundamental that the B-notions. In
principle the B-notions have to be defined in terms of the A-
notions. Even the notion of temporal possibility can be
defined on terms of the A-notions.
Priors idea of instant propositions:

Hybrid logic.

Instant propositions: a, b, c .

(I1) a: a
(I2) ~~a
(I3) (a p) (a ~p)

We can define T(a,p) as well as the before-after

relation, <:
(DT) T(a, p) def (a p)
(DB) a < b def (a Fb)

Priors view on time and logic:

Time is not an object, but

whatever is real exists and acts
in time... But this earlier-later
calculus is only a convenient but
indirect way of expressing truths
that are not really about events
but about things ... [undated
note]
Priors ideas of instant propositions and branching time:

An instant should be understood as a very rich proposition

which implies everything which is true.
In this way the whole branching time system is included in
the instant. This means that any part of the system contains
the whole system.

Now

The A-language and branching time

Prior made no use of Peirces Gamma Graphs. This is

strange since it seems attractive from an A-theoretic
point of view.

Challenge:
How can Priors tempo-modal systems be formulated in
terms of Peirces Gamma Graphs?
 the question as to whether
there are or could be
unconnected time-series is a
senseless one. but these
diagrams cannot represent
time, as they cannot be
translated into the basic non-
figurative temporal language.
[Prior 1967, p.199]

4a 4b
1 2 3

6
 4b 4a 3 2 1 5 6

4a 4b
1 2 3

4b 4a 3 2 1 5 6

4a 4b
1 2 3

4a 4b 6 3 2 1 5