Intellectual Practices of Keith Lyons
A platform can disappear while the work on it still matters.
In April 2014, SlideShare was about to discontinue its audio-track service. Keith Lyons had used the platform to share presentations with accompanying audio. The presentations would remain, but after the end of the month the audio would not.
Keith did not simply record the loss.
He began curating the audio tracks on Clyde Street. Where one recording exceeded WordPress’s file-size limit, he pointed to a copy in his public Dropbox folder. He left the presentations on SlideShare, considered what audio contributed to asynchronous sharing, and began testing Prezi and Evernote as alternative presentation environments.
The public work had already been published. The platform changed anyway.
For the whole-corpus study, that raised a more difficult question than whether Keith shared things openly:
What had to happen before a piece of public work counted as deliberately kept alive rather than simply followed by something new?
A public object can have a declared future
The earliest approved case makes the operation unusually visible.
In August 2008, Keith created a WordPress introduction connected to a keynote he was preparing for the 2009 International Association for Sports Information World Congress.
He wrote that, although he had first posted the introduction on 12 August, “my introduction to you here will continue to change.”
Then the post actually did.
The source contains dated additions stretching across the following months: Posterous and Google Video experiments, Iterasi archiving, a first CCK08 post, Facebook and Netvibes links, a PBwiki project, a Ning site, Technorati, a conference abstract in Scribd, audio versions of the abstract and, by March 2009, a draft SlideShare presentation with a SlideCast.
This matters because the evidence is not merely an intention to update something later. The existing public object visibly acquires dated additions and new forms.
Publication is therefore not being treated as a seal placed on the object. The page has a continuing public history.
But this does not mean that every long page, amended post or changing website qualified. The whole-corpus boundary was stricter. Public work had to be deliberately maintained, extended, revised, adapted or explicitly left open to further development. Length, age, availability and serial posting were not enough by themselves.
A draft can be public before the event it is for
A different form appears in May 2010.
Keith was preparing for a panel at the Australian Institute of Sport’s Performance Research Workshop. The panel question was how research effort in high-performance sport might be made more effective.
Before the discussion took place, Keith published his presentation on Clyde Street and described it plainly: “This is a draft of my thoughts hosted by SlideShare.”
He had already added a slide making a case for the Australian research network to work together as an open-access community. He also said he was looking forward to hearing the views of the other panellists and exploring possible synergies with Gavin Reynolds’s account of the National Sport Information Centre.
The important move is not that a presentation was available online. It is that Keith exposed it as a draft before the encounter for which it was being made.
That leaves the object publicly capable of change before the live discussion has even happened.
The source does not show that the panel subsequently changed the presentation. It does not justify a claim that public access improved the discussion. It does not show audience uptake.
What it does show is enough for the narrower finding: the public artefact is not presented as final simply because other people can already see it.
A revisable public object does not have to be one page
The strongest cases also prevented the study from reducing revisability to editing a document in place.
In December 2011, Keith opened a football post with a simple statement: “I have been monitoring discussions about goal-line technology in football.”
He referred back to his report from the previous month, shifted from Australian Rules Football to Association Football, added new information about testing and governance, and recorded that a decision by the International Football Association Board was still pending for March 2012.
Here the continuing public object is a monitoring record distributed across posts.
That distinction matters. A series can become a living public record when later entries deliberately extend the same monitored object under new developments. But seriality alone does not establish that operation.
This case was also an important boundary against a different kind of return. Keith was continuing the public monitoring record as new external developments arrived. He was not materially reworking an earlier intellectual object into a new form. The whole-corpus study therefore retained this as revisable public work rather than treating every return to an older topic as a deeper reanimation of earlier work.
Comments can extend a post without transferring ownership
Revisability sometimes appears in much smaller moves.
In May 2016, Keith wrote Getting it wrong, reflecting on an AFL prediction that had failed badly and using the match to think about decision support for coaches.
The original post is already substantial. But at the end Keith added a marked postscript:
“A comment by Graham on this post has led me to Ryan Buckland’s discussions of AFL in The Roar.”
The post has acquired a new route after publication.
Ownership remains important. Graham owns his comment. Ryan Buckland owns his analysis. The study did not absorb either person’s intellectual work into Keith’s.
Keith’s operation was narrower: he visibly extended his own public post after later input led him somewhere else.
Nor does the source establish that Graham’s comment changed Keith’s underlying AFL analysis. The evidence supports an extension of the public object and a new source route, not a stronger claim about audience-caused revision of the inquiry itself.
Unfinishedness can be an operating condition
By April 2017, Clyde Street contains a particularly explicit case.
Keith had been reviewing research on shots and goals in ice hockey and association football. He compiled a bibliography covering 2004 to 2017 and made his developing work available through Google Docs.
He wrote:
“My paper is incomplete but extends to twelve pages. It is a Google Doc so I will continue to update it.”
He also described his literature synthesis as an ongoing Google Doc project, created a partial open educational resource about football analytics, and said he was keen to develop the resources and would welcome guidance.
This is a strong case because unfinishedness is not merely visible. It is connected to an explicit decision about the public object’s continuing status.
The documents are incomplete and public, and Keith said they would continue to be updated.
The source also contains a different kind of incompleteness: some ice-hockey articles could not be accessed. Keith recorded the missing access and kept those references in the list. That limitation belongs to the evidential boundary of the literature review.
The revisability finding concerns something else. It concerns the decision to keep the bibliography, synthesis and educational resource open as developing public objects.
The distinction matters because unfinishedness by itself did not qualify. A draft abandoned in public, a broken link, an incomplete series or a page that simply stops are not evidence of deliberate revisability merely because they remain unfinished.
Another output is not necessarily another life
The negative cases were essential because Clyde Street contains a great deal of serial analysis.
In November 2013, after the first Ashes Test at the Gabba, Keith published a short post titled GPS: Gabba Partnership Skyline 3.
The post states that the Test had concluded and presents the partnership skylines for the two teams.
It is public. It belongs to an ongoing cricket-analysis programme. It contains another visual output.
But this source does not show an existing public object being deliberately maintained, revised, adapted or left open. It mainly presents another result from an established series.
The whole-corpus study retained this kind of record outside the finding.
That control prevents revisability from becoming a synonym for persistence.
Keith could keep analysing a subject without keeping the previous public object live. He could publish another graphic without revising the earlier graphic. A season could continue. A tournament could produce another instalment. A new result could be added to a familiar sequence.
All of that may be worthwhile.
None of it automatically shows that the public work itself has been given a continuing life.
The public object can survive a platform without becoming a different inquiry
The 2014 Slidecasts case makes another boundary visible.
SlideShare’s withdrawal of audio tracks was a materially changed technical condition. Keith responded by preserving the audio elsewhere, keeping the visual presentations available and testing other platforms.
But the earlier intellectual content of the presentations was not substantially reworked in the source.
That is why the case belongs here rather than being treated as evidence that changed conditions necessarily transformed the earlier inquiry.
The public objects were being preserved and adapted so they could continue to exist publicly despite the platform change. Their survival under platform change is the operation.
This is a useful distinction in a digital archive. Technical migration can be intellectually consequential without automatically becoming a new intellectual argument.
A late reflection makes the public orientation explicit
In November 2019, Keith returned directly to blogging as an object of reflection.
He discussed Stephen Downes, Ben Werdmuller, academic citation of blogs and social media, the tension between fast-moving online conversation and conventional publication, and the longevity of public research.
Then he located his own practice inside that discussion:
“I have always seen my blog posts as my thinking out aloud”.
He described moderating comments before publication and reported almost 2,000 conversations about Clyde Street posts.
This late source is useful because it makes Keith’s own account of public writing more explicit.
But it must not be read backwards as proof that every earlier Clyde Street post was deliberately revisable, or that one stable philosophy governed twelve years of publishing.
The whole-corpus finding does not depend on that retrospective statement. It depends on the source-level cases across the chronology where public work is actually maintained, extended, adapted, corrected or explicitly left open.
What the whole corpus established
Across the frozen 2,051-record Clyde Street corpus, 95 records were retained as positive evidence for this operation.
Seventy-seven were primary decisions. Eleven earlier multi-code records contained the operation, and seven later cases retained it as a secondary co-occurrence.
The earliest approved positive is from 12 August 2008. The latest is from 5 February 2020. Approved positives occur in every calendar year from 2008 through 2020 and across thirty-one chronological batches.
The evidential core is strong but not uniform. Sixty-nine cases have finding-specific Explicit strength. Nine primary cases remain Probable at the lower boundary. Ten late primary cases are approved positives but do not retain separately recoverable finding-specific strength or boundary detail, and seven later secondary co-occurrences likewise lack a separate finding-specific strength assessment.
Those provenance gaps are preserved rather than filled from inference.
The resulting judgement was high confidence for recurrence, temporal spread, ownership and the strict revisable-public-object boundary. Fine-grained developmental description was held slightly lower because the corpus contains lower-boundary serial cases, late detail gaps and long stretches without an approved positive.
The forms diversify across the chronology: dated additions to posts, public drafts, cumulative monitoring records, open course and project objects, presentation preservation and adaptation, postscripts after comments, OER revision, ongoing Google Docs, work-in-progress presentations and explicit reflection on blogging as public conversation.
That diversity does not warrant a story of increasing sophistication, importance or commitment over time.
Later cases often make the mechanism easier to locate. The study did not convert that greater visibility into a developmental claim.
What did not count
The negative space is large.
Ninety-two explicitly routed rivals and near-misses were retained to constrain the finding. Sixty-one are explicit countercase or no-code decisions; the remainder are cases where another operation more cleanly explains what is happening. That portfolio is not a denominator of every record that failed to qualify. It is a boundary set showing how easily publicness, updating and recurrence can be mistaken for deliberate revisability.
The dominant exclusions are clear:
another instalment in a series without demonstrable maintenance of the public object or record;
a refreshed graphic;
a year-end summary;
format conversion without demonstrated continuing development;
routine correction or platform housekeeping;
public availability without active maintenance;
prospective intention to revise something later;
and substantive development owned by a collaborator rather than Keith.
The finding also remains separate from neighbouring ones.
A public object can be revisable without exposing enough of its construction for another person to inspect how it was made.
A public object can be useful to others without being kept open to revision.
A limitation can constrain a claim without making the public document a living object.
And changed conditions can cause an earlier work to be materially reworked in ways that go beyond ordinary continuation or preservation.
Keeping those distinctions separate was necessary to prevent a broad idea such as “open work” from swallowing several different intellectual operations.
What the finding does not permit
The whole-corpus evidence does not permit a claim that all blogging is revisable work.
It does not permit a claim that every update, version, postscript, serial result or platform migration qualifies.
It does not show that readers noticed the revisions, used them, benefited from them or caused them.
It does not show that leaving work open is better than stabilising it.
It does not establish that Keith was uniquely committed to revisability as a personal trait.
And it does not justify turning the cases into a universal method in which all public work should remain indefinitely unfinished.
The bounded finding is more specific.
Across Clyde Street, Keith repeatedly treated some public intellectual objects as capable of further life after first release. He kept them live through extension, revision, adaptation, correction, cumulative public development or explicit openness to further work.
Just as importantly, the study repeatedly refused to call ordinary publishing continuity the same thing.
What the distinction leaves with the reader
The easy lesson would be: keep everything open.
The evidence does not support that.
Some objects need stability. Some require version control. Some should be superseded rather than silently altered. Some are useful precisely because a reader can rely on a settled version. Others gain value from continuing development.
Clyde Street does not supply one rule for choosing among those possibilities.
What it makes visible is the choice itself.
Publishing something and deciding its future relationship to change are not always the same decision.
The question the finding leaves behind is therefore:
Question to carry
Which pieces of my public work am I deliberately leaving capable of another life — and what would I need to see before I could say that life actually continued?
Sources followed
Keith Lyons, “Slidecasts” — 14/04/24
Keith Lyons, “IASI in Canberra 2009” — 08/08/12
Keith Lyons, “HPRW10 Connected” — 10/05/13
Keith Lyons, “Goal Line Technology Update December 2011” — 11/12/07
Keith Lyons, “Getting it wrong” — 16/05/09
Keith Lyons, “Shots and Goals: Quality, Expectations and Context” — 17/04/20
Keith Lyons, “GPS: Gabba Partnership Skyline 3” — 13/11/24
Keith Lyons, “Blogging” — 19/11/16