I have just come across an interesting blog and wiki by Brian Lennon, who discusses ConnectedText in the context of academic note-taking. I would warn people who are new to wikis and ConnectedText to keep in mind that Brian is dealing with fairly advanced features in CT, such as using Python scripts. You don’t need to understand any of this stuff and still be able to use CT successfully as a note-taking tool and much more. Nevertheless, Brian’s use of CT is a good example of its versatility and expandability with scripts and plugins, which can lead to integration with other tools, such as Zotero.
In my previous post I discussed how I turned to writing as a way to put a stop to my obsessive-compulsive outlining behaviour. Outlining is certainly an important part of preparing for writing, but not if outlines keep on begetting further outlines and there is no end in sight to the proliferation of outlines. The white space of the empty Word document acted as a laboratory or workshop where I could cobble together the various conceptual threads, while still keeping a reverse outline to monitor the emerging shape of the argument. This strategy worked fairly well for a while, and I ended up with about 3000 words that I’m reasonably happy with.
Then I got stuck. Now, I’m sure lots of things have been written about the writer’s block, and I don’t have time to read any of it. So apologies if I’m stating the obvious. But I’ve been finding that there are usually two reasons why I get stuck: either I haven’t worked out the content of what I’m going to say (the conceptual point) or I haven’t worked out in sufficient detail the practical steps (workflow) that I need to take. The two reasons are closely interrelated because usually I need to work out the practical steps to be able to develop the conceptual points.
At this stage of the game however I couldn’t fathom developing yet another linear outline in a traditional one-pane or two-pane outliner to work out the next steps. One reason why I’m finding linear outlines less helpful towards the end of my project is that this stage requires synthesis, which is about putting things back together, rather than laying them out in a line (out-line). Synthesis requires matrix-type tools, as there are at least two “dimensions” that need to be brought together (if we consider a linear outliner a one-dimensional tool).
There are a number of ways to create such a matrix effect. One could just use a table (Word) or a spreadsheet (Excel) or columns in a one- (Bonsai) or two-pane (myInfo) outliner. I particularly like to use TreeSheets for this type of matrix-outlining. However, this time I reached for my favourite concept mapper, VUE, again, as it was a lopsided matrix that I had in mind, where one axis had a lot more content than the other and I needed the ability to visualise the shape of the whole thing, no matter what size it grew to.
I am working on a literature review chapter and I needed to evaluate the work of a major author. I had to review my notes on seven of his works (kept in ConnectedText), extract the main points and organise them into separate outlines in Bonsai, and finally consolidate them into a single outline. It was this last step that I used VUE for. I ended up with a concept map that looks like this:
The matrix is constituted by the analytical lens on the right-hand side (grey nodes), which was applied to evaluating the rest of the nodes. When I started the mapping and review process, I thought I only had three sub-topics to write about. Conceptual development occurred as part of this mapping process, and I ended up discovering that I had three more important things to say. Finally at the bottom I drew some conclusions from the material.
The advantage of doing this in VUE (as opposed to a traditional one- or two-pane outliner) was that I didn’t need to worry too much about an overall hierarchy initially. Any node can be linked to any other node, and while there is a certain hierarchy here (the overall argument flows from top to bottom, and section arguments flow from left to right and down), working in a concept map feels more fluid and free than in a linear outline. It is just a matter of different visualisation, as all of this could have been done in an adequately equipped one-pane outliner as well.
So what are the next steps? Now that I know what I want to say in this part of my chapter, I can write an introduction that will state what is going to be said, with a thesis sentence that combines the topic sentences of the six sections and points towards the contributions that will be summarised in the conclusions. My plan is to keep this concept map open in my second monitor and use it as a guide while I dictate my text into Word with Dragon.
I am in the middle of writing my dissertation, and I’m constantly experimenting with my writing process flow and set-up. I have two monitors, a 19 and a 22-inch one. The overall writing process appears to be a recursive process of alternating between outlining and writing (though the two are not always clearly separable).
My actual writing started once I got to a point that I had so many outlines in different outliners that the only way out of that morass was to start writing them up. Let’s call that condition obsessive-compulsive outlining behaviour (OCOB), and the cure “writing.” The reason for the many outlines is that I use outliners as part of a distillation process, for extracting and abstracting information by translating ideas from one outliner into another. This is a fun and productive process until it gets out of hand and you end up with more outlines than you can keep track of. At that point I decided to start writing, so that the written piece be the place where all the information from the various outlines can be consolidated into a final train of thought.
To my surprise I have selected Word 2010 as my main writing tool. I considered Scrivener and DragonPad, but I found that Word trumped them by being compatible with the most software that I wanted to use during the writing process. For one, I could dictate directly into Word and all the Dragon NaturallySpeaking functions would work the same as in DragonPad, making DragonPad redundant. Scrivener on the other hand couldn’t handle direct Dragon dictation. But I also already had a detailed outline in Scrivener (using it as a two-pane outliner), so it seemed sensible to keep that open in my second monitor, while doing the writing in another software in my main monitor.
For times when I wouldn’t be using Dragon, I would want to use WordExpander to speed up the typing, and it turned out that it would display the pop-up menu of word choices by the cursor in Word but not in Scrivener. Finally, Word has an EndNote add-in, so I could directly be inserting references as I write, rather than using a more roundabout way of referencing in Scrivener by having to paste in raw EndNote code first, and then having to convert it once the Scrivener draft had been exported into Word. While Scrivener has strengths when it comes to tracking word count and to switching from an individual section view to a total document (Scrivenings) view, word count in Word is acceptable and I can use headings to navigate the document when the Navigation Pane is enabled.
So, Word is open in my 19-in monitor that is facing me and where I do the final writing up (dictating with Dragon or typing with WordExpander). I also have a little 4-inch (10 cm) of a NoteTab Light window open on the right of the Word window, where I type quick notes I need to refer to or text fragments I want to use a bit later.
In my 22-in monitor that I have on my right-hand side at an angle, I have a number of software windows open, though not all at the same time. At most there would be two open at any one time, arranged vertically.These would include my three main outlines which I have given up on consolidating into a single overall outline. They are in Natara Bonsai (a single-pane outliner), Outline 4D (a single-pane outliner with inline notes), and Scrivener (a two-pane outliner). Scrivener holds my overall outline for the whole PhD, for all 8 chapters. Outline 4D holds a detailed outline just for my current chapter.
The advantage of Outline 4D over Scrivener (as an outliner) is that it is single-pane, so I can see the entire text with its headings and sub-headings in a single window. Although Scrivener does have the “Scrivenings” view which combines all its documents into a single view, it is more of a ‘tape’ view than an outline view. Plus in Outline 4D you can select different colours and fonts for the different hierarchical levels, which makes navigating and comprehending the outline easier.
Finally, I have an even more detailed outline of just one section of the chapter in Bonsai. Bonsai’s disadvantage over Outline 4D is that it does not have inline notes (notes can be displayed in a separate pane to the right of the outline or at the bottom, which is just not the same). On the other hand Bonsai is faster to operate than Outline 4D, can cram more information into the same amount of window space, and it also allows you to colour in hierarchical levels.
Besides the outlines, I also need to have ConnectedText (CT) open, as it contains my reading notes. This information is more detailed than the one in the outlines. If I need even more detail about the particular source (usually an academic article), then I look at the corresponding PDF with the original article in PDF XChange Viewer (it’s called up by one click on the PDF link in CT). Finally, I also have Directory Opus open, should I need to look for a file that way; EndNote, so I can insert references directly into Word as I’m writing; and Waterfox, so I can look up things on the Internet, when necessary.
This describes the set-up of my outline-to-writing process. It’s a matter of gathering and weaving the information from the various application windows together into the written text in Word. However, the reason I called this “recursive outlining and writing” is because after I had written a reasonably big junk of text, I need to go back and outline some more. The reason for this is that the actual written text has its implicit outline of its own, which does not fully correspond to any of the original outlines. First, it already started out as an amalgam of the other three outlines, so it naturally can’t be identical with any of them perfectly. However, the work of outlining and conceptual development continues during the writing process itself. The writing itself is the workshop where some of the final ideas get eventually hammered out. The text has some new content.
In order to fully grasp the implicit structure of the argument that I have developed in the course of the writing, I need to re-create it in another outline. (It’s difficult to discern this structure easily in the written text itself, once you have thousands of words in a complex academic-type writing). Let’s call this step “retrospective” or “reverse outlining.” The perfect tool for this I found in VUE. I did try Noteliner first, but I found that using a visual concept-map worked better than yet another text-based outline.
I maximised the VUE window to the entire 22-in monitor, and starting at the top, flowing from left to right, and then downwards, I started to map out the core argument of each paragraph in the form of a concept map. When the screen filled up, I just continued scrolling down (there seems to be unlimited space). I also used colour to mark out important moments such as theses, propositions or key findings. Each major section of the chapter is represented by nodes in different colour. Arrows are used to represent logical relationships.
This is the overall shape of the concept map for the chapter that I’m currently working on, in a 25% zoomed-out view (text becomes visible at a 50% zoom):
This reverse visual outline helped me find gaps in my argument, as well as redundant sections. It also worked as an analytical tool, as it helped me discover new relationships and come up with new findings and interpretations. It also helps with developing a sense for what to write about next. This recursive process of outlining-writing-outlining (with the help of VUE as the virtual whiteboard where I sort all my conceptual issues out) is likely to become a permanent part of my writing practice. I am even considering printing these VUE maps out and pinning them on the wall. This would enable me to survey the entire line of argument from chapter 1 to chapter 8 of my dissertation, that represents a text potentially the size of 80,000-100,000 words.
Surfulater is half-price today on BitsduJour. I’m happy to put in a good word, as Surfulater has been a crucial part of my setup for many years. It is my main database for storing entire webpages or snippets of text and images from webpages. I like to keep my collected web pages separate from my other types of collected digital data, which I organise in ConnectedText, as I just find it helpful to keep those two worlds separate. It’s easier to find stored webpages this way, as Surfulater has a hierarchical tree-based organisation that can be searched very quickly with the Filter search bar. If you do want to link to a specific item in Surfulater’s tree from an external application (such as ConnectedText), that is also possible.
If you do qualitative research, you will quickly discover that websites can not only change frequently but they also disappear without warning, so it is critical to capture data from them periodically with a tool like Surfulater. I use Surfulater mostly with Firefox (and more recently with Waterfox) and occasionally with IE. It ‘s as easy to use as 1-2-3:
- Highlight the information you want,
- right-click and
- select “Surfulater: Add new article” or “Surfulater: Attach page to article” (which will save the entire page as it is).
And that’s it. The information will be automatically saved in a folder in Surfulater’s hierarchical tree.
Dragon NaturallySpeaking 12 is an indispensable tool for me and this weekend (begins midnight Thursday, ends midnight Monday) Nuance has a 50% off sale for the home edition, and also for some other software, at its UK online store. I’m not sure if these offers are also available to non-UK shoppers but it might be worth a try. Here are the prices:
Dragon NaturallySpeaking 12 Home – £79.99 £39.99
PDF Converter Professional 8 – £99.99 £49.99
PaperPort 14 – £49.99 £24.99
OmniPage 18 – £79.99 £39.99
I already have DNS 12, but I’m looking for an OCR application to convert entire (including scanned) PDF articles to text (that would be OmniPage then), so I can import them into ConnectedText in one go and then delete the unnecessary bits, rather than continuing with my current piecemeal method of converting bits of text with the otherwise excellent ABBYY Screenshot Reader.
After a bit more research I’ve decided to pay the extra couple of pounds and go for ABBYY FineReader 11 Pro. For one, it has generally better reviews than OmniPage 18, but also I’ve been so impressed with their Screenshot Reader – which I got for free and have been using for years – that I’m happy to reward them for that.
Update no. 2:
Wow, OCR applications have moved on since the last time I used them a few years ago! So far I’m very pleased with ABBYY FineReader. It only took a couple of seconds to convert a 17-page PDF article to Word, complete with not only all the footnotes and headings but also all of the images.
The main purpose of the scanning is to end up with a text version to be imported into ConnectedText. The scan wasn’t entirely perfect in the sense that the first page with the abstract ended up appended to the end of the Word document, but it was easy to fix that with a cut and paste job.
There were a couple of italicised words in the original that weren’t kept in italics in the scan and some block quotes didn’t end up looking the way they were supposed to, however the OCR-ing, which matters to me the most, was very close to 100% accuracy.
Update no. 3:
Correction: in fact the ABBYY FineReader scan was perfect: the problem I mentioned regarding lost italicisation and indentation of block quotes had nothing to do with ABBYY. Those features got lost in the conversion process from .docx to .htm and the CT import process. Apologies to the ABBYY folks…
I have added a Replaced page to the blog, which will document my justification for removing tools from my toolbox and replacing them with other tools. So I have a bit of a system going here. The Favourite tools list describes the contents of the current toolbox, the Watchlist has a list of candidates under consideration for adoption into the toolbox, and the Replaced list has the tools that have been removed from the toolbox.
I created a new page called Watchlist, which will contain a regularly updated list of links to software, hardware and other solutions that I may try out or revisit in the future.
Here is a mini case study on how to organise your physical library of books with the help of Story Turbo (v. 2.2), a virtual corkboard software. (There is also a slightly cheaper version called Story Lite without the image support.) My particular purpose for this exercise was to come up with a project plan for reviewing literature for my dissertation, for which I have only about 9 days available. I have of course analysed my literatures before, but now I need to pull everything together for one last time for writing up my literature review chapter.
Just to be clear, here I’m only talking about organising and analysing an actual library of around 200 physical books which have taken up about five shelves in my bookcase. I need to scan them visually for relevant quotes that I may use in my dissertation. The organisational challenge is to identify exactly which books are the important ones and what literatures are exactly represented on my bookshelves.
(Obviously if I had been using ConnectedText to keep track of my hard copy book readings from the start, I wouldn’t have to be doing any of this. Fortunately my reading notes of electronic articles have been kept in better order in a WhizFolders database. But there is no doubt in my mind that I will be using ConnectedText for all my new reading notes for the foreseeable future.)
STEP 1: Organise your books on the top of a large desk (or on the floor) into thematic groups (distinct areas of literatures). In my case this has resulted in 7 overall groupings, within which there were up to four sub-groups. I organised these piles of books in “columns” and “rows”, imitating the format of a table or spreadsheet in a software.
STEP 2: Open Story Turbo (or Story Lite) on your desktop and represent each pile of books with an index card (I made mine look like yellow post-it notes on a grey background). There was a one-for-one correspondence between a pile of books and an index card, the overall effect being a series of columns and rows of yellow post-it notes on a virtual corkboard in the exact same order as my book piles. I ended up with twenty index cards in Story Turbo, organised into seven columns.
STEP 3: Analyse, organise and annotate your library. You may discover new relationships between the groupings (areas and sub-areas of literatures) and may want to re-organise both the virtual index cards and the corresponding physical piles of books to reflect the new order. You can add as much detail as you like to each index card. If a pile of books is too big and too diverse (e.g. 20 books with 5 sub-themes), then you could create additional index cards and break up the pile into five smaller piles. The goal here is to organise your library and also capture electronically both the overall structure of the library and the descriptions of its content.
STEP 4: One of the main reasons for using Story Turbo is to be able to export the outcome of this analysis in a number of ways. I have chosen an RTF export, as I wanted to manipulate this list further.
STEP 5: I copied the contents of the RTF export from Word and pasted it into Natara Bonsai. It wasn’t the cleanest of exports, in the sense that the order of the list didn’t perfectly reflect the order of index cards in the columns, but it was only one or two items that needed to be repositioned out of 20. Also, the RTF contained some horizontal lines and line breaks which needed to be deleted in Bonsai. However, doing this sort of list editing is a breeze in Bonsai. I reorganised my list into a hierarchical outline where the content was grouped thematically. After a bit more analysis I had realised that the seven larger groups could be consolidated into five groups.
STEP 6: In order to come up with my project plan to review these books, I allocated days according to the importance of each theme to the research question and the number of books available in each pile.
STEP 7: I added the allocated themes (representing piles of books that are still on my desk) as all-day appointments to my Google Calendar, to serve as a reminder, so I know how much time I have available for reviewing each pile. All I need to do know is start working through the piles of books each day, and slowly start clearing my desk as I put the processed books back on the shelf. (By “processing” I mean reviewing my handwritten comments and post-it notes in the books and recording the selected quotes and notes in my ConnectedText database – more on which later).