Keyword "tagging" of text passages

rv
rvdparis
Posts: 64
Joined: Fri Mar 09, 2007 12:47 am
Location: France

Fri Aug 19, 2011 2:23 pm Post

I would like to ask for this feature in a future version of Scrivener: Highlight a passage of text (in a text document, not necessarily in a PDF document, though that would be nice too), add a Keyword to it.

I have brought it up before here: http://www.literatureandlatte.com/forum/viewtopic.php?f=4&t=7206&p=58529&hilit=keyword#p58529 and here: http://www.literatureandlatte.com/forum/viewtopic.php?f=2&t=14652

My argument for doing so is:

1. It would allow quick searching and finding of textual passages that are relevant to a specific keyword. This feature would allow several keywords per document permitting a writer to quickly find, break up and rearrange ideas within a paragraph or document.

2. It would allow qualitative social science academics who work with texts--interviews, letters, memoranda, observational notes, survey responses, etc.--to analyze them with keywords and then search for them later during writing up. A process called "coding." Why Scrivener? Because the process of analyzing those texts is also a process of writing up your findings. It would be great to have the texts of my interviews in my Research folder, with Keywords marking up my interviewees' responses, and then finding them while I am writing my draft.

Possible issues:

    Overlapping keyword passages, where you tag a passage with keyword A, and you tag part of that passage with another keyword, B.

    The database of keywords and locations may be too large (a document might have hundreds of keywords)

This is a wish for the wish list; I realize fully that Scrivener is NOT a qualitative data analysis package, and in fact, the program is really for other kinds of writing, but every time I use Scrivener, I just have this background voice that keeps saying, "this program is so well suited to academic work, and if it just made some tweaks, it could work for social sciences too."

In any case, thanks for an amazing program that I use on a daily basis. Rich

User avatar
Jenny_Y8S
Posts: 189
Joined: Tue Jun 01, 2010 7:49 pm
Platform: Mac
Location: Surrey

Sun Aug 21, 2011 8:21 pm Post

As it's likely this feature isn't going to be implemented (at all? / soon?). How about this?

Highlight the appropriate block of text and add a comment to it. In your comment, type in your "Keywords". This way your comment becomes the repository for the keywords pertaining to the appropriate text.

You can build collections baed on searches for the keywords and clicking on the comments in the inspector will take you right to the marked text.

Yes this has numerous limitations, eg it's always visible, you have to manage your own keywords, you can't have overlapping text segments etc. Some of them you can get used to, other's you can mitigate against the negatives by using project replace etc

But...

It's not 100% ideal, but it works, and the features to implement it exist already.
Feed me chocolate, I'm developing Scrivener for iOS.

User avatar
Jenny_Y8S
Posts: 189
Joined: Tue Jun 01, 2010 7:49 pm
Platform: Mac
Location: Surrey

Sun Aug 21, 2011 8:22 pm Post

Another approach is of course to split the documents into sub documents within one folder. You can still view, edit and compile the document as one but you can now assign your own meta data to each component part.
Feed me chocolate, I'm developing Scrivener for iOS.

rv
rvdparis
Posts: 64
Joined: Fri Mar 09, 2007 12:47 am
Location: France

Mon Aug 22, 2011 10:40 am Post

Those are all good solutions.

What bugs me is that instead of searching for text in comments or footnotes in the search field, you have to go to Edit | Find... | Find by Formatting, which I feel is a bit awkward and counter-intuitive (I am SEARCHING for text, not looking for it by how it has been FORMATTED),

AND you can't save your search.

Or I don't know what I am talking about. How do you "build collections based on searches" that you do among the Comments and Footnotes sections?

Having said that, I am hopeful that the developers can reflect on this issue since I do think it would be valuable to a lot of people. R

User avatar
KB
Site Admin
Posts: 20775
Joined: Tue Jun 13, 2006 11:23 pm
Platform: Mac
Location: Truro, Cornwall
Contact:

Mon Aug 22, 2011 11:23 am Post

...I am hopeful that the developers can reflect on this issue...


You mean developer (singular) - me. :)

As I said in the other thread you started about this, there are no plans for anything like this currently, sorry. I'll bear it in mind when I eventually come to start mapping out features for 3.0, but I can't promise anything, as personally I think tagging individual paragraphs and sentences - whilst useful for some users I am sure - is overkill in Scrivener.

Thanks for the suggestion and kind words though.

All the best,
Keith
"You can't waltz in here, use my toaster, and start spouting universal truths without qualification."

User avatar
AmberV
Posts: 24311
Joined: Sun Jun 18, 2006 4:30 am
Platform: Mac + Linux
Location: Ourense, Galiza
Contact:

Mon Aug 22, 2011 4:38 pm Post

rvdparis wrote:Or I don't know what I am talking about. How do you "build collections based on searches" that you do among the Comments and Footnotes sections?


Project Search by "All" does go through comments and footnotes as well, and Project Searches can be saved using the magnifying glass into a Collection tab. What you need to do, to make this technique more effective, is codify your keywords if they are like normal words, and then include the codification as part of the search. For example, if I have a note attached to some text because it requires further research, I can add a comment to that range, and then type in "RSRCH //" I might put some thoughts after the slashes if there is something to illuminate in the original text that might not be obvious in the future; that part is optional. Now I can use project search to look for that string, save it to a tab, and always have every document that needs further research collected for me. When I am done with the comment, I can either remove it or change the codification so that it no longer matches---to a "done" state you could say, like "RSRCH #". Naturally, this technique works for inline annotations as well (which is what I prefer for this sort of thing).

The main disadvantage to this technique is that it does not build anywhere a list of used terms. It is a solution that takes advantage of string searching; not a dedicate feature like Keywords. You probably wouldn't want to have hundreds of saved search tabs, and setting one up each time you use a new code (after checking to make sure you haven't already done so) would be labour intensive.

Perhaps AppleScript in the future will help with this task. If you could run a script that looked for " //" in a comment and then added a keyword for whatever string precedes the match, creating one if necessary, then you could run this AppleScript periodically when it came time to start consulting lists.
.:.
Ioa Petra'ka
“Whole sight, or all the rest is desolation.” —John Fowles

rv
rvdparis
Posts: 64
Joined: Fri Mar 09, 2007 12:47 am
Location: France

Mon Aug 22, 2011 5:01 pm Post

Thank you Keith for considering this option in a future version of Scrivener. Honestly, if I could recreate the scenario I want--"tagging" a text passage, giving it a term, and then compiling the terms later in a search--then I would leave you alone, promise :)

So AmberV's suggestion seems very suitable. I didn't know Comments were included in Project Searching, and now that I can save that kind of search, I am even happier. My basic need was really to identify all the passages in my book where I refer to a chapter or section so that if things change in that chapter or section, I can find them all easily and update them to reflect my change. Now I will comment those passages with "Chapter:X", do and save a search, so that in the end, I'll have one search collection per chapter. I tend to not use the search feature too often, so this solution won't be too intrusive.

I will content myself with using Keywords for documents and subdocuments. Which is what Jenny_Y8S was hinting at above; that I could create subdocuments for each paragraph and have those keywords (or comments) at my command.

Thanks again for your input and help, Rich

rv
rvdparis
Posts: 64
Joined: Fri Mar 09, 2007 12:47 am
Location: France

Tue Aug 23, 2011 10:00 am Post

I can't seem to find how to change the default yellow color of the comments background neither in the editor nor in the side bar where the comments are kept. Is this possible? (I'm sure I'll figure it out before someone responds!)

UPDATE: I knew i'd find it. You have to ctrl-click on the comments to change their colors. Is there a default setting in the Preferences I am missing?

User avatar
KB
Site Admin
Posts: 20775
Joined: Tue Jun 13, 2006 11:23 pm
Platform: Mac
Location: Truro, Cornwall
Contact:

Tue Aug 23, 2011 1:48 pm Post

Hi,

No, there's no default setting - once you change the colour, though, the latest colour will be used by default until you change it again.

All the best,
Keith
"You can't waltz in here, use my toaster, and start spouting universal truths without qualification."

Et
EthnoManc
Posts: 3
Joined: Tue Aug 23, 2011 2:17 pm
Platform: Mac

Tue Aug 23, 2011 8:47 pm Post

Hi there,

Since having moved from a PC to a Mac I too have been searching for something to replace NVivo, which I have used a lot in the past, but to be honest never been totally happy with - clunky user interface and slow when working with a lot of data. As an ethnographer I basically need precisely the functionality mentioned in this thread - as I read through my fieldnotes I need to tag chunks of text with keywords and at a later date to gather together all the pieces of text linked to any particular code. I'm always amazed how difficult it is to tag fieldnotes using normal writing software. Anyway I have just tried out Jenny_y8s' 2nd suggestion of splitting the document into subdocuments and can report that the results are superb - I haven't found anything that comes close to this on a mac, which is remarkable given that this is not what Scrivener is even designed to do. Whilst it doesn't quite have the flexibility of being able to have overlapping codes for different chunks of text it looks like this goes a long way towards emergent and intuitive coding and subsequent analysis of textual documents, which is great.

It would, as Rich suggests, be fantastic if in a future version it were possible to do this without splitting up the document. I've only been using scrivener for a short while but I am incredibly impressed with the way in which it can assist in writing academic papers, and this functionality would make this an even more outstanding piece of writing software for qualitative researchers.

Just a quick final question - what is the best way of creating a sub-document from a paragraph/sentence of a document - when I split the document it just creates another document at the same level, rather than making it a child of the original. Is there a better way to do this?

Hannah

User avatar
AmberV
Posts: 24311
Joined: Sun Jun 18, 2006 4:30 am
Platform: Mac + Linux
Location: Ourense, Galiza
Contact:

Tue Aug 23, 2011 8:59 pm Post

Just a quick final question - what is the best way of creating a sub-document from a paragraph/sentence of a document - when I split the document it just creates another document at the same level, rather than making it a child of the original. Is there a better way to do this?


Not immediately, but once you've split the document off and named it, use Ctrl-Cmd-RightArrow to demote it. Or, if you prefer the mouse, just drag & drop it beneath the item it came from. Once you have done this once, any further splits to the parent will automatically end up as a child for that node. Splits within that set will of course be siblings unless you alter them.
.:.
Ioa Petra'ka
“Whole sight, or all the rest is desolation.” —John Fowles

vi
vijen
Posts: 1
Joined: Thu Sep 08, 2011 10:00 am
Platform: Mac

Thu Sep 08, 2011 10:13 am Post

Hello!

I've been using Scrivener for a few days now and it has already capsized the way my writing works. It feels much more natural which helps when tapping on creativity.
I write on this post, since I've been desperately looking for the same functionality mentioned here: I would like to tag parts of my text and then be able to search through those tags later on. This should be done through a shortcut to avoid having to stop the writing flow. This is particularly important when reporting and taking notes in a journalistic manner where notes have to be taken quickly and many ideas come up to mind while writing.. these ideas for instance ("Read More", "Questions to Ask") are embedded in the text and therefore some kind of natural mechanism should exist.
I am not aware of any other software more suited to this kind of process than Scrivener. A qualitative analysis software is not helpful for instance in journalism or note taking.
Personally, I think the the keyword functionality is not as useful as this "text-tagging".. this is because, keywords is not a natural process since you have to stop your flow to think about key terms that identify your text. I am not good at using keywords and it feels like the extra time doesn't give me that much of a return. Perhaps text-tagging could be the evolution of the keywords.

Anyways, these are just some reflections. Please keep us updated about this.
Keep rocking
Cheers

User avatar
KB
Site Admin
Posts: 20775
Joined: Tue Jun 13, 2006 11:23 pm
Platform: Mac
Location: Truro, Cornwall
Contact:

Thu Sep 08, 2011 11:10 am Post

Hi vijen,

Why don't you use comments or inline annotations for this? From your description of what you want, these things should be perfect for that sort of use. You can then use the formatting finder (Edit > Find > Find by Formatting) to quickly cycle through all comments or annotations, even searching for particular text in comments or annotations (e.g. "Read More" or whatever). And the comments or annotations can be excluded from the final text.

Hope that helps.

All the best,
Keith
"You can't waltz in here, use my toaster, and start spouting universal truths without qualification."

Je
JennyD
Posts: 2
Joined: Sat Oct 29, 2011 12:37 pm
Platform: Mac

Sat Oct 29, 2011 1:57 pm Post

Hello!

I can sense you might be getting a bit tired of these requests now but I'd just like to add my vote of support for possible-future keyword "tagging" of text passages.

Can I just say first that I - like everyone else, it seems - am quite in love with scrivener, and not at all keen to transfer over to Nvivo or any of the over horrible clunky qualitative data analysis softwares out there. I'm an anthropologist and I don't think it's a coincidence that I've spotted a few more of us in here, making similar queries. Unlike other social scientists, we don't really have any need for the many fancy bells and whistles that dedicated QDA softwares offer; only the ability to navigate our way through vast tracts of fieldnotes. And for this, it's true, the ability to highlight and code text passages would be incredibly useful.

The alternatives you've suggested sound like they might work fine but do keep it in mind for future versions of scrivener. Since you are marketing the software to researchers now as well as novelists - you might be surprised how popular this simple feature would prove (not only with anthropologists but also, I imagine, historians, literature students, and people working throughout the humanities).

Thanks very much - keep up the good work!

User avatar
AmberV
Posts: 24311
Joined: Sun Jun 18, 2006 4:30 am
Platform: Mac + Linux
Location: Ourense, Galiza
Contact:

Sat Oct 29, 2011 9:12 pm Post

The alternatives you've suggested sound like they might work fine but do keep it in mind for future versions of scrivener.


What would be helpful is to define what is merely "fine" about the current features; or in other words, where to they break down and cease being useful, to merit only being "fine". What about them does not qualify them as "the ability to highlight and code text passages". To my mind, Scrivener has about a half a dozen different broad categories by which you can do that, and probably dozens of possible use-cases involving the variations of using them and/or combining them into compound tools.

Keep in mind Scrivener's toolkit platform based philosophy, rather than having "features", just like novelists don't have character tracking and plot tension features. They instead have the same tools you can bend toward QDA---but they are using them to track where a particular detail of the story is located in the text. So where do these broadly useful tools not work.
.:.
Ioa Petra'ka
“Whole sight, or all the rest is desolation.” —John Fowles