avatar image

Welcome, Guest. Please login or register.
Did you miss your activation email?

Login with username, password and session length
  • Tuesday March 31, 2020, 12:24 pm
  • Proudly celebrating 15 years online.
  • Donate now to become a lifetime supporting member of the site and get a non-expiring license key for all of our programs.
  • donate

Author Topic: Retrospective Outlining  (Read 5175 times)

Paul Keith

  • Member
  • Joined in 2008
  • **
  • Posts: 1,987
    • View Profile
    • Donate to Member
Retrospective Outlining
« on: May 25, 2012, 06:50 PM »
Copied from: http://www.outliners...rospective-outlining

This is interesting more from the way it is presented than the actual content. At least to me, it just seems like typical outline draft editing. For the most part, I prefer to just quote Stephen Zeoli whose opinion I most agree with it (though not all of it):

This is an interesting topic. It relates to a feeling I’ve long had about outliners, namely that the ones which of restrict your outline to individual blocks of text are not really outliners at all. They are hierarchical information managers. A true outliner allows you to see the entire structure—including the text related to each heading of the outline—in a single window so that you not only understand the structure, but you also understand the flow of words.

In other words, what you’re looking for is an outline view and a document view, in which the document view shows your entire project in a single pane. Changes made in either view are instantly reflected in the other. Alternatively, an application could work the way old GrandView did. It was a single-pane outliner in which your text was visible inline, but could be switched on or off, so you could see it all, or just the structure. In my view, none of the current outliners (OmniOutliner, Neo) handle inline text well enough to really pull this off.

In fact, the number of applications which actually do this is very small. Scrivener comes close with its scrivenings view, which shows your project in one flowing document, and has the option of displaying titles. You can edit titles and text in the scrivenings view and those changes are reflected in the Binder. However, in a quick test I just made it looks like you can’t rearrange the sections in the scrivenings view without messing up the Binder. (This relates to the Mac version of Scrivener. I doubt the PC version is even this sophisticated.)

I would say that ConnectedText is likely the best at this, another reason for admiring it.

Steve Z.

Programs mentioned:

ConnectedText -

sc10.pngRetrospective Outlining

Writer's Block -

wb3boxshot.jpgRetrospective Outlining

What is retrospective outlining:

Rebecca Nowacek: If a regular outline is something you write before you draft out your paper, a reverse outline is something you do after you write a draft.

Dr. Andus:

I could call it “retrospective” or “reverse” outlining. Usually when we talk about outlining we mean a process that precedes writing or writing-up. However, I find that in the process of writing the structure of my text often changes and at the end it departs significantly from the original outline. Then I need to engage in “retrospective outlining,” i.e. drawing up an outline that reflects the new hierarchical logical structure of the now completed draft. The purpose of this is to get an overview of the argument and structure of a large document (10k+ word social science paper or book chapter).

In the past I just used Natara Bonsai and later on Noteliner to reconstruct the implicit outline. Then I discovered MS Word 2010’s navigation pane, which can display a retrospective outline (Table of Contents) if you apply headings. But I was never fully satisfied with that solution, as visually the hierarchy is not that clear and it takes up a lot of screen space to display complex hierarchies, not to mention that I may not want my document to have so many headings. So it’s a trade-off between having a ridiculous amount of headings or not having a detailed enough retrospective outline.

And this is where ConnectedText comes in. I have just realised that CT’s Table of Contents pane in fact can be used as a real-time outliner (i.e. it doesn’t even have to be retrospective). As I’m writing a text and creating hierarchical sub-headings, all I need to do is have the TOC open somewhere (usually docked on the left) and switch between edit and view modes for the real-time outline to be displayed instantly. This is extremely useful, as I no longer need to wonder about the nature of the evolving logical structure while I’m writing or have to reconstruct it afterwards.

Why should I reverse outline?

The reverse outline can be an extremely useful tool for helping you see the big picture of your paper, and can be especially useful for papers in need of major reordering of paragraphs or papers filled with paragraphs that have too many ideas in them and therefore don't hold together.

The Brainstorm Way:

Alexander Deliyannis:

For me outlining has always been a continuous two-way process in the way you describe and back again; I suspect for others here as well. The tools I use for writing support this way of working, though clearly some are better than others. I would note two aspects to what you describe:

(a) Bottom-up development of the outline; focus on the detail writing and the outline will build/update itself.
(b) Text re-organisation: I expect that when one looks at an outline developed from (a), s/he will become aware of possible weaknesses, e.g. over-emphasis on one argument and under-representation of another, in which case the top-down process would recommence, by adding/removing headings etc.

The way I see it, Connected Text seems indeed very well suited for (a); as Steve noted recently , you can start by creating ‘cards’ and think about the structure later, as connections between the cards develop. I know of many ‘visual’ tools that can do this, as well as (b), but most would be useless for texts growing to the thousands of words with tens of headings.

So, with this application in mind:

- For (a) I can think of two very powerful tools, namely Brainstorm and Sense; a separate post should follow sometime soon from my part on Sense, which is developing very nicely.

- For (b) I would think again of Brainstorm --JB has built a significant part of his Cyborganize system on Brainstorm’s powerful re-organisation features- Sense, as well as MaxThink.

So it’s a trade-off between having a ridiculous amount of headings or not having a detailed enough retrospective outline.
-Dr Andus wrote:

I expect that in Word you can set up styles that are identical in everything except for their outline level. So, for example, below level 3 subheading you could have several more paragraph levels that would be properly indented in the outline view, but without appearing as headings in the text. The disadvantage is that you would have whole paragraphs in the outline.

For such ‘micro-outlining’ I find that Sense is ideal; in the outline you can go down to paragraph level if you want, effectively having a bird’s eye view of the full text.

The Limitations of ConnectText:

Dr. Andus:

I think CT is not quite there yet either. The problem at the moment is that if you are working on a long document, then the table of contents view gets long too, and so switching back and forth between the view and edit mode requires a lot of scrolling in the TOC pane. Although you can collapse headings and thus make the TOC text appear without the scroll bar, for some reason CT expands the collapsed headings every time one saves the document or switches between the view and edit mode (switching is required for updating the TOC).

Alexander, thanks for the SENSE suggestion. I looked at it in the past and I couldn’t quite figure out what I could use it for. But if it can do this kind of iterative, retrospective outlining with large documents, then that would be an interesting niche and worth taking another look.

I will check out Writer’s Blocks too, though the price is a big disincentive for spending time with it, given the features that it seems to have, which seem rather basic. But that’s just my first impression and perhaps an unfair one.

As for Brainstorm, I have the same problem as I had with CT for many years. Whenever I looked at it I just couldn’t get my head around it quickly enough to carry on. But as my CT experience had just taught me, there might be rewards for persevering…

Edwardo Mauro:

If you have any suggestion of how we can improve CT regarding this, let me know.

BTW, we just released a new version.

Dr. Andus:

Hi Eduardo,
I don’t know how easy it is to do technically but if the collapsed headings would stay collapsed when switching from the edit mode to the view mode (and thus updating/refreshing the TOC) that would do the trick. Then one could work on a very long document because by collapsing level one headers for instance the TOC text would be visible in the pane and there would be no need to scroll.

An alternative and improved solution could be to have a keyboard shortcut or a button somewhere that could update the TOC without having to switch from edit to view mode (and also keeping collapsed headers collapsed).

Thanks for your consideration. I’m looking forward to upgrading to the new version.

How ConnectText fixed the issue with real time changes:

Well, I have just trialled the forthcoming version of CT (I think it will be and Eduardo had not only resolved the above issue but he turned the Table of Contents (TOC) pane into a proper real-time, live outliner. This means that as you type in the edit window, the TOC now automatically displays the developing outline (the hierarchical structure of headings 5 level deep). No need to switch from “edit” to “view” (or hit “save“) in order to update the TOC. The outline (or parts of it) will now stay collapsed or expanded, regardless of the length of the outline. Brilliant!!

The added bonus for me is that I also use the TOC as a qualitative analysis tool for annotating long texts (a process similar to “reverse outlining“), and that now happens live too.

So if I said that “CT was not quite there yet,” now it’s definitely there and probably beyond :) Thanks Eduardo!

Analyzing the Contents of a Reverse Outline:

Go through the paper and number each paragraph. Then on a separate sheet of paper, write #1 and the main point (or points) of that first paragraph. Then, on the next line write #2 and the main point(s) of the second paragraph. Go through the entire paper this way. When you have gone through the entire paper, you will have an outline giving you an overview of your entire paper.

Then what?

    * Now look carefully at your overview, asking yourself the following questions:

    * Are the paragraphs properly focused, or are there multiple main ideas competing for control of a single paragraph?

    * Now that you've identified the main point of each paragraph, does the topic sentence reflect that point?

    * Are some of those ideas in a paragraph extraneous and should they therefore be deleted from the paper? Or do they simply need to be moved to a different part of the paper? (Many times you may find that a random idea tacked onto the end of, say, paragraph five really belongs in paragraph eleven where you fully develop that idea.) When you look at the outline as a whole, does the organization of the paper reflect what you promised in your introduction / thesis? If the answer is no, consider whether you need to revise the thesis or revise the organization of the paper.

Paul Keith

  • Member
  • Joined in 2008
  • **
  • Posts: 1,987
    • View Profile
    • Donate to Member
Re: Retrospective Outlining
« Reply #1 on: May 25, 2012, 06:59 PM »
One important criticism I have with the feature fix is that it goes against the spirit of "reverse" outlining but it is as hollow as saying "if you don't follow GTD to the letter, you can't make GTD to do work".

The spirit of reverse outlining is that you have to make a report on your own finished work and by adapting real time changes, it motivates one to get lazy and just outline on the go.

However at the same time, real time is real time. It is a convenience on it's own and pure reverse outlining is not some panacea. It could basically be summed up as a more formalized manner of reviewing your own work with as much integrity as when you typed out your original work. 

Another failure on my part is that I haven't tried the latest version of ConnectText and am just going by with what was said.

The key problem with reverse outlining will always be the same though. Software outline design tends to fix the logging convenience. Reverse outlining as a method is not about fixing the logging but enabling a more correct and in-depth capture of the things you missed therefore any type of "pane", "preview". "notes box", etc. only works if it assumes the original work is close to well done or the actual wielder of the tool is great at scrutinizing his own work. If they aren't, real time IMO kills the retrospection.