Some of Tufte's criticism of slide bullets is not specific to PowerPoint, but could be extended to software which work with nested nodes like Roam/Workflowy(software which I like). Workflowy, in fact, allows you to convert a tree of nodes into a linear presentation. Wittgenstein also wrote his classic books as numbered nodes (though not nested).
Indeed, this can also be seen as a critique of structured code editors vs text editing. Mathematics books also follow tree structure a bit, (Def 6.3.1, Example 6.3.2), though there often is some connecting narrative.
The point about oversimplification to fit into a single slide is specific to PowerPoint. But, the critique that organization into discrete nodes often skips over an underlying narrative or a causal structure which connects the nodes is more general.
What can be said in defense of discrete organization? Firstly, the overall narrative is not initially apparent. Listing the pieces together can help to discover this structure.
Secondly, in long essays, the larger point often gets buried in the details. This is especially true in mathematical works where the purpose of a complicated definition/result is only seen a long time later. This also happens in source code, where a lot of preprocessing obscures the central purpose of a function (though of course, source code is not a candidate for a report with sentences anyway).
By forcing these documents to become less dense, the narrative actually becomes more apparent. Whereas with a dense document, the reader's attention can wander away before the punchline.
One issue that Tufte seems to not discuss in the oversimplification critique is that attention/time is limited. Since an organization leader cant read all reports 3 levels down the ladder (either usual style reports or nested trees), there needs to be a strategy for marking specific reports as important and also to mark which details from the document need to be passed on to higher levels of decision making and which details should be only relevant to middle managers.
In the Columbia report, the problem is not oversimplification but that the critical conclusion was mentioned as a low level detail whereas a methodology of choosing a 'conservative' model became the heading.
Could a usual technical report have avoided this issue? The 'conservative' phrase could well have been a section heading and the damage indicated in sentence buried inside the section. But a technical report also has a 'Conclusion' section which could have forced the authors to state their position clearly. This 'Conclusion' section is implicitly a protocol for which information in the report has to be passed up to a higher level. IPCC Reports have a 'Summary for Policymakers' in discrete points (https://www.ipcc.ch/sr15/chapter/spm/). Tufte, for some reason, doesn't like "Executive Summary".
I've worked on heavy PP based companies, and in companies that requires 5 minute 2 pagers.
Both are here to write (info dense) but having a proper narrative structure with the 2 pager communicates far more than the PowerPoint does. A narrative structure can also drop down to bullet points it needed.
I've seen plenty of 50 page PP decks that ended up with slides full of paragraphs, basically being the worst of all possible communication formats.
Indeed, this can also be seen as a critique of structured code editors vs text editing. Mathematics books also follow tree structure a bit, (Def 6.3.1, Example 6.3.2), though there often is some connecting narrative.
The point about oversimplification to fit into a single slide is specific to PowerPoint. But, the critique that organization into discrete nodes often skips over an underlying narrative or a causal structure which connects the nodes is more general.
What can be said in defense of discrete organization? Firstly, the overall narrative is not initially apparent. Listing the pieces together can help to discover this structure.
Secondly, in long essays, the larger point often gets buried in the details. This is especially true in mathematical works where the purpose of a complicated definition/result is only seen a long time later. This also happens in source code, where a lot of preprocessing obscures the central purpose of a function (though of course, source code is not a candidate for a report with sentences anyway).
By forcing these documents to become less dense, the narrative actually becomes more apparent. Whereas with a dense document, the reader's attention can wander away before the punchline.
One issue that Tufte seems to not discuss in the oversimplification critique is that attention/time is limited. Since an organization leader cant read all reports 3 levels down the ladder (either usual style reports or nested trees), there needs to be a strategy for marking specific reports as important and also to mark which details from the document need to be passed on to higher levels of decision making and which details should be only relevant to middle managers.
In the Columbia report, the problem is not oversimplification but that the critical conclusion was mentioned as a low level detail whereas a methodology of choosing a 'conservative' model became the heading.
Could a usual technical report have avoided this issue? The 'conservative' phrase could well have been a section heading and the damage indicated in sentence buried inside the section. But a technical report also has a 'Conclusion' section which could have forced the authors to state their position clearly. This 'Conclusion' section is implicitly a protocol for which information in the report has to be passed up to a higher level. IPCC Reports have a 'Summary for Policymakers' in discrete points (https://www.ipcc.ch/sr15/chapter/spm/). Tufte, for some reason, doesn't like "Executive Summary".