A Possible Future in the Frame — Beyond the Breadcrumbs

In June 2009, Keith Lyons returned to an idea that had been with him for a long time.

He was writing about Peter Dowrick’s work on self-modelling. Keith remembered encountering it while he was researching video in sport in the 1980s. What caught him was a possibility that made video do something very different from simply replaying an error.

A learner might see themselves “not so much as they were but as they might be”.

The words belonged to Dowrick’s line of research. Keith did not claim them as his own. What Clyde Street lets us follow is what he did after keeping them.

That distinction matters because Peter Dowrick appears differently from many people in Clyde Street. There is no need to manufacture a personal relationship or a story of one decisive encounter. The source pass did not locate evidence here of correspondence, collaboration or friendship between them.

What is visible instead is an intellectual source that Keith kept returning to.

Over the next decade, feedforward moved through video, learning, open courses, coach development, performance analysis, mental time travel, data sharing and artificial intelligence. The word remained recognisable. The work Keith asked it to do did not.

That is the case worth following.

More than replay

Keith’s 2009 post, Feedforward, reached back before Clyde Street.

He reconstructed a trail through Dowrick’s work on video self-modelling, including early work in swimming and later applications in disability, reading and other settings. Keith’s interest was not simply that a recording could show somebody what had happened. He was drawn to the possibility of constructing a view of successful behaviour that the learner had not yet fully performed.

That altered the temporal direction of the recording.

Feedback could look backwards.

Feedforward could make a possible next performance available for inspection.

Keith connected that possibility with questions he had developed through coaching, guided discovery and work on learning. This was already a re-use rather than a repetition. The research remained Dowrick’s. Keith’s contribution was to bring it into his own problem: what might video become if it helped a learner recognise possibility rather than merely revisit deficiency?

It is tempting to turn that into the whole story.

Clyde Street does not let us.

By 2011, a talk by Royce Sadler had sent Keith back towards Dowrick, Ian Franks and Gary Miller, Kristine Chambers and Joan Vickers, Richard Schmidt and others. The immediate problem was no longer self-modelling alone. Keith was thinking about learner flourishing, the timing and quantity of feedback, and how much intellectual work a coach should leave with the learner.

Peter remained in the room, but he did not own the room.

That is an important boundary for the case. The long lineage is visible precisely because Keith kept placing Dowrick beside other people rather than treating one source as a complete answer.

The idea changes rooms

In 2012, feedforward appeared inside a different working environment.

Keith was running #OAPS101, an open course in sport informatics and analytics. Conversations around the course had encouraged him to think about performance analysts as educators and educational technologists.

Here the old self-modelling formulation became a new question.

What if performance analysts shared not only what had happened, but something of the world as it might be?

The movement is small enough to miss. A concept developed through video self-modelling was now being asked to help rethink the purpose of an analyst’s output.

The analyst was not only a recorder of the past.

The output could become part of somebody else’s learning.

That early part of the feedforward story already appears elsewhere in the Clyde Street publication corpus, particularly in the learning encounter Who learns from the analysis? The Peter Dowrick case earns its separate place only because the trail does not stop there.

Two years later Keith made the return unusually visible.

In October 2014, while preparing another open course, #UCSIA15, he explicitly went back to the feedforward post he had written in 2009. He was now reading Dowrick’s more recent work on mental simulation, mirror neurons and mental time travel. The possibility of learning from a constructed future event widened the older video question.

Keith wrote that he was fascinated by the idea that we could learn from the future.

The changed setting mattered. This was no longer just a coach and athlete watching selected video. Keith was designing a public learning environment in which participants could move through material asynchronously, connect their own experiences and follow different routes through the course.

Dowrick’s research remained Dowrick’s. The open-course architecture was Keith’s.

The connection Keith made between the older source and the course became a new Clyde Street object.

Same word, different meanings

The most revealing return came in October 2016 because Keith encountered another use of the same word.

Stephen Downes was talking about feedforward in the context of ongoing professional development and changing what we know. Keith did not smooth the two meanings together.

He distinguished them.

Downes’s use of feedforward was not simply Dowrick’s self-modelling account. Keith then made the difference productive. He placed Stephen’s formulation beside what he called his long-term interest in Peter Dowrick’s work, an interest he located back in the early 1980s.

That source is unusually useful because it shows both continuity and change in the same place.

Keith had not forgotten where his own meaning had come from.

But he was also willing to let another meaning put pressure on it.

He connected Dowrick, Downes and further writing about surfing, balance and change to think about continuing professional learning. A formulation originally discussed in relation to children learning to read could now help him consider adults learning in changing professional environments.

That transfer is Keith’s.

It should not be mistaken for a finding Dowrick had established in coach education.

The source is valuable for exactly that reason. It lets us see intellectual inheritance without intellectual surrender. Keith could preserve the provenance of an idea and still rework what it was doing.

The previous day, the same lineage had entered practice more directly.

Keith was preparing to work with rugby coaches. He described himself as feeding forward to the session and drew on the possibility of learners seeing successes they had not yet had.

That is a realised design decision: the idea entered the preparation of a real learning encounter.

It is not evidence that the coaches learned what Keith hoped they would learn.

Clyde Street gives us the design, not the outcome.

From memory to possible futures

By September 2017, the problem had shifted again.

Keith was thinking about what performance analysis itself was for.

For much of its history, video and analysis had helped augment memory. A coach could revisit events that had happened too quickly, too far away or too densely to remember accurately.

Keith now asked whether the service could augment the future as well as the past.

Dowrick reappeared explicitly. Keith described his feedforward work as seminal and connected it with a much wider literature on theory of mind and mental time travel.

This is where a simple influence story begins to fail.

The later argument did not belong to Peter alone. Keith was reading across neurocognitive work on imagining future events, episodic memory and simulation. Other researchers were doing substantial intellectual work inside the post.

What Keith owned was the reframing.

He wanted performance analysis to become a co-produced service in which people could use evidence not only to account for what had happened but to consider what might happen next.

He wrote that if the service embraced feedforward — “as I think it must” — there was more work to do around future-oriented simulation and decision making.

The phrase is stronger than a passing citation.

It is still not proof that Dowrick caused the reframing.

A more careful reading is better: an old intellectual source remained available when a new professional problem appeared, and Keith combined it with newer evidence to alter the question he was asking about analysis.

When the source becomes less visible

Long intellectual lineages do not remain equally explicit.

That is part of this story too.

In January 2017, Keith had written to the Secret Soccer Analyst and included a question that had become important to him:

How might feedforward transform our practice?

Peter was not named in that post.

In October 2018, while thinking about momentum, prospection and learning design, Keith wrote that newer work spoke to his fascination with feedforward. Peter survived there almost as a piece of metadata — a tooltip attached to the term — while the immediate intellectual work was being done by other researchers.

On the last day of 2018, Keith found a paper about unlearning while searching for recent discussions of feedforward. That post connected feedforward self-modelling with visualisation, accessibility and Keith’s wish to become a better data sharer and storyteller.

Again, Peter was not the principal visible source.

These are not failed examples.

They protect the case from becoming too neat.

An idea can remain in somebody’s intellectual vocabulary after its original source stops being named every time. Other people can extend it, alter it or independently use the same word. New problems can make different literatures more important.

We should therefore resist saying that every later appearance of feedforward is evidence of Dowrick’s continuing influence.

Sometimes the lineage is explicit.

Sometimes it is only background.

Sometimes the word is doing work that cannot responsibly be assigned to Peter at all.

And then, in September 2019, Peter returns to the foreground.

Keith was thinking about the verbs used to describe performance analysis in a digital era.

He remembered an older professional vocabulary:

Observe.

Record.

Analyse.

Model.

The environment had changed. Data science, connected practice and artificial intelligence were part of the changed setting Keith was trying to describe. He placed those older verbs beside IBM’s AI Ladder and asked what was missing.

His addition was short:

“I would add to this feedforward.”

Then he returned to Dowrick’s work on mental simulation and possible future events.

The late significance of Peter in Clyde Street is not that an idea from decades earlier had predicted artificial intelligence.

The sources do not support that claim.

The interesting thing is that Keith still found the old source useful when the surrounding technology changed dramatically.

Feedforward had travelled from edited video towards a verb for professional practice.

That is a different job.

Keeping the source and the reworking visible

The Peter Dowrick trail is therefore not a straight line.

It does not show one theory moving unchanged from the 1980s to artificial intelligence.

Nor does it show Peter quietly authoring Keith’s later work from a distance.

It shows something more modest and, I think, more useful.

Keith found an idea early in his work with video. He remembered where it came from. He returned to it publicly. He put it beside other sources. He let new problems change what it was being asked to do.

Sometimes the source was named prominently.

Sometimes feedforward had become part of Keith’s own working language.

Sometimes other researchers carried the inquiry further.

The continuity lies in a recurring possibility: analysis need not be confined to telling somebody what has already happened.

The transformations lie in what Keith built around that possibility.

A learner might see a performance not yet achieved.

An analyst might design information for learning rather than delivery.

An open course might let people explore possible futures.

A coach-learning session might be organised prospectively.

A performance-analysis service might help people imagine what comes next.

A profession encountering artificial intelligence might need a future-facing verb.

None of those later moves belongs to Peter simply because his work sits somewhere in their ancestry.

But leaving the ancestry visible matters.

It lets us distinguish the intellectual gift from the work done with it.

That may be the strongest reason for keeping this trail in Clyde Street.

Not because it gives us a model to copy.

Because it gives us a question about our own intellectual inheritances.

Which ideas have we carried for years?

Can we still say where they came from?

Question to carry

And when we use them in a new setting, can we show what belongs to the source — and what we have done since?

Source and ownership note

Peter Dowrick and his coauthors retain ownership of their research, formulations and findings; other cited researchers retain ownership of theirs. Keith owns the way he returned to, connected, re-situated and applied those sources in his own public inquiry and practice. No direct Peter–Keith correspondence, collaboration or personal relationship was established in the source pass, and repeated return is evidence of durable intellectual use rather than proof of causality.