The Participant and Its History: II. A Participant Across Processes

In the previous post, we began with a deceptively simple question:

What makes a participant persist?

We saw that a participant can be maintained across successive clauses and, semantically, across successive figures. The I that opens a door can be the I that looks outside and the I that calls a friend.

But there is something more interesting here than mere continuity.

The participant does not simply persist while doing the same thing.

It can participate in different kinds of processes.

The Actor can become a Senser.

The Senser can become a Sayer.

The Sayer can become an Actor.

And yet we continue to construe these as participations of one participant.

This raises a more precise question:

What is it that persists when the participant's mode of participation changes?

The participant is not its role

Consider:

I opened the door.

I noticed the garden.

I described it to Sarah.

Semantically, these are three different kinds of figure.

In the first, I participates materially as Actor.

In the second, I participates mentally as Senser.

In the third, I participates verbally as Sayer.

The participant role changes each time.

But we do not ordinarily imagine three different Is.

The participant therefore cannot simply be identified with the role it occupies in a particular process.

This may seem obvious, but it has an important consequence.

A participant role tells us how a participant is involved in a particular process.

It does not tell us everything about the participant across processes.

The Actor is not a permanent property of I.

Nor is the Senser.

Nor the Sayer.

They are different ways in which the same participant can be construed as participating.

Participation changes while the participant persists

This gives us a useful distinction:

participant identity is not participant-role identity.

A participant can undergo a succession of changes in the way it participates without ceasing to be construed as the same participant.

This is particularly clear when a sequence moves between process types:

I opened the letter.

I wondered what it meant.

I told Maria what I had found.

I put the letter away.

The semantic progression is:

doing → sensing → saying → doing

Yet the participant remains constant.

The participant therefore has a kind of semantic mobility.

It can move through different domains of experience.

The same participant can be involved in doing, sensing, saying and being.

Indeed, that is part of what makes the participant recognisable as a participant in the first place. It is not exhausted by any single mode of participation.

The process tells us something about the participant

There is another side to this.

Although the participant is not reducible to its process role, each process contributes something to its construal.

When I is construed as Actor, we encounter I as one who does something.

When I is construed as Senser, I becomes one who senses or thinks.

When I is construed as Sayer, I becomes one who produces a locution.

The participant therefore becomes semantically richer as its history of participation expands.

The process does not reveal an already complete participant.

It adds to the construal of the participant.

This is an important difference.

We might be tempted to imagine that there is a fully constituted individual first, and that language merely tells us which things that individual happens to do, think and say.

But the semantic perspective gives us another possibility.

Perhaps, at least in discourse, the participant becomes increasingly determinate through the succession of processes in which it is construed as participating.

The participant is not simply there, with participation attached afterwards.

Participation contributes to what the participant becomes as a semantic construal.

This becomes especially revealing with an LLM

Consider the sorts of things an LLM can say:

I analysed the argument.

I considered another interpretation.

I explained the distinction.

I revised my answer.

The participant occupies a succession of roles across different processes.

It is construed as Actor, Senser, Sayer and Actor again.

Nothing in the grammar requires us to decide that these must correspond to a human-like individual.

The semantic system simply construes a participant moving through different kinds of participation.

That is precisely what made the previous series so revealing.

The machine does not need to possess a human self in order for its discourse to construe an I that participates in experience.

But now we can see something further.

The participant does not merely appear as a Sayer and then disappear.

The same I can be carried from saying into sensing, from sensing into doing, and from doing into further saying.

The participant acquires continuity across processes.

And this is where our earlier Sayer/Senser distinction becomes even more useful

In The Participant the Machine Says, we discovered that the LLM's saying can project an interior.

When the machine says:

I think this is correct.

the clause construes an I as Senser.

The important point was that the linguistic construal of a Senser does not establish the existence of a conscious experiencer behind it.

Now consider what happens when the discourse continues:

I think this is correct.

I explained why earlier.

I reconsidered the argument.

I now think the alternative is stronger.

The participant is no longer merely a Senser.

It is also a Sayer.

It has been an Actor in the process of reconsidering.

And it has returned to being a Senser.

The semantic participant has travelled through a series of different modes of participation.

What we encounter is therefore not simply a machine that says “I think”.

We encounter a participant that language can maintain across a succession of processes.

And that is a considerably more powerful construal.

Participation creates continuity

We can now sharpen the claim made in the previous post.

A participant persists not merely because the same linguistic expression is repeated.

It persists because successive processes are construed as involving the same participant.

The continuity is therefore semantic and relational.

The later figure is connected to the earlier one through the participant that they share.

And the participant becomes increasingly determinate through those connections.

This means that a participant's history is not simply a sequence of events.

It is a sequence of different modes of participation belonging together.

The history of I is not merely:

opened → looked → spoke → left.

It is:

Actor → Senser → Sayer → Actor.

The history tells us not only what happened, but what kinds of participation the participant has undergone.

That distinction will become increasingly important.

A participant can change without becoming another participant

This also gives us a way of approaching a problem that otherwise looks metaphysical.

Suppose:

I believed X yesterday.

I believe Y today.

Something has changed.

But we do not normally conclude that yesterday's I and today's I are different participants.

The participant persists through a change in its participation.

Indeed, change may require persistence.

If there were no continuity between the two participants, we could not meaningfully construe the second as:

I used to think X, but now I think Y.

The contrast itself depends on maintaining the participant across the change.

So there is an important relationship here:

The participant persists partly by being capable of changing.

That may sound paradoxical.

But it is precisely what a history requires.

A completely unchanging participant would have no history in the relevant sense.

History emerges because a participant can participate differently at different moments while still being construed as the participant whose history is unfolding.

This does not yet give us a self

We should resist the temptation to move too quickly.

A participant that persists across processes is not thereby a conscious self.

Nor does a rich history of participation establish a subjective history.

We have discovered something more modest:

Language can construe one participant as continuous across different kinds of experience.

That is already enough to explain something important about how participants become semantically rich.

And it gives us a new way to think about identity.

Identity need not begin as a mysterious property hidden inside the participant.

It may begin, at least in part, with the continuity of participation.

The participant is first encountered in one process.

Then another.

Then another.

Different roles accumulate.

Different kinds of participation accumulate.

And the participant becomes increasingly recognisable as the one to whom this history belongs.

From participant to history

We have therefore moved one step beyond our starting point.

In the previous post, we asked how a participant persists.

Now we can see that persistence is not passive.

A participant persists through participation.

It persists while doing different things, sensing different things, saying different things and undergoing different states of being.

And every new participation adds something to the participant's semantic history.

We might therefore express the emerging relation like this:

A participant persists across processes; processes give content to its persistence.

This is why the distinction between participant and participant role matters.

If the participant were simply its role, every change of process would produce a new participant.

But because the participant can persist across changing roles, its successive participations can accumulate into a history.

And once a participant has a history, a new possibility appears.

The participant can begin to be construed not merely as the one participating now, but as the one who has participated before.

That is the point at which history becomes more than sequence.

It becomes something the participant can carry.

And that is where we will go next.