Humans + AI Remi Kalir on social annotation, self curation, the connective tissue of ideas, annotation tools, and nuance for synthesis (Ep31)
“What is nuance? I think it’s really important, particularly in today’s information environment, whether we’re talking about the challenges of accurate information, conspiracy theory, or in general trying to find the signal through the noise.”
– Remi Kalir
About Remi Kalir
Remi is Associate Professor of Learning Design and Technology at the University of Colorado Denver, and a leading scholar of annotation. He is co-author of the book Annotation, published by MIT Press, and many journal articles on the subject. Remi is also Scholar in Residence at Hypothesis and the co-founder of Marginal Syllabus.
What you will learn
- What is a scholar of annotation (01:44)
- What is social annotation (03:52)
- What is the difference between highlighting and annotating (07:13)
- How to get started in annotating (10:03)
- How to build associative trails across multiple texts (12:51)
- Why social annotation is useful in collaborative or collective learning (21:00)
- What visual tools for social annotation are available (22:07)
- Why we should pay attention to nuance for annotation and idea synthesis (25:30)
- Why reading slowly, with other people, and a robust annotation practice is a great set of practices to thrive on overload (30:13)
Episode resources
Transcript
Remi: Ross, thank you for the invitation. I really appreciate it.
Ross: You are a scholar of annotation. I want to know more about that. What is an annotation?
Remi: Let me begin by saying Ross that you’re an annotator. You are an annotator; I know that without even knowing the nuances of your practice but I know that you read. I know that when you read, you probably make your thinking visible. I know that when you read, you take notes. It’s very likely that when you were in school, or for pleasure on a couch, reading poetry, wherever it may be, you’re probably marking up a book, or you’re probably taking notes in some kind of way. If you’re listening to this podcast, you also probably have a personal history with annotation.
One of the things that I find fascinating about the practice of annotation is that people are annotators. When people read, they are writing; when they are reading, they are thinking, they are making connections. That has to be made visible in some way so that gets to that core definition of what is annotation. It’s the addition of a note to a text. It’s that simple. You add a note to a text. We might do that when we read books.
Again, for so many of us, we’ve done that in formal educational settings, schools, and universities, maybe when we were very young, maybe when we were older. So many of us have personal histories of adding notes to texts, of being annotators. Of course, we also do that in much more sophisticated ways now. We do that in digital environments. We do that with all kinds of fancy online tools. But at the very core, at the basic definition of that, we’re adding notes to texts, as we engage with media, as we engage with information, as we try and navigate information overload. I know you’re an annotator Ross, I’m an annotator too, and if you’re listening to this, you’re probably also an annotator. I find that to be very fascinating.
Ross: We’ll definitely want to dig into that from all sorts of angles. But one point is you also talk about social annotation, we’d love to hear about that.
Remi: That’s an important distinction, and thank you for making that, which is is that for many of us, let’s say we are reading a book of poetry, or maybe we’re reading something online, we can mark that up, we can add our notes to that text in a private way. Much marginalia, if you want to use that term, a lot of book marginalia is private, it’s written only for us, it’s written for an audience of one. Yet, we can also make our annotation social, we can share it, we can use, again, either a book lent to a friend, or we can use a whole variety of digital and online tools to make our marginal notes accessible to other people.
There’s a whole genre now of not only technologies but also practices that allow us to share our thinking with other people. We can do so to appreciate different perspectives, we can do so to disagree, we can do so to build consensus, but social annotation is a practice, it’s an opportunity to make those cognitive processes, to make those social processes more connected, as we read together, as we think together, as we make sense of the world together. Again, another reason why I find social annotation a very promising practice, both in and outside of schools, and in other kinds of learning institutions.
Ross: One particular example of that is that you are co-author of the book “Annotation”. I believe the process of writing the book involved social annotation.
Remi: It did, we had to walk the talk, so to speak. When my co-author, dear friend, and my colleague, Antero Garcia, from Stanford University, when we wrote the book, we came up with our draft… we’re not thinking alone, we’re not doing this work in isolation, so we wrote up a draft of the entire book and we used a particularly agile, intuitive platform that comes from MIT, our publisher. It’s called the PubPub platform. We put the whole book there. We said, hey, smart people, thoughtful people, people who disagree with us, hey, come read this draft of our book.
This social annotation process can be very nuanced. We can find a word or a phrase or a particular argument, we can highlight it, we can add commentary, we can reply. We generated literally something like 20,000 words worth the word count volume of the annotations. It was an incredible corpus of rich feedback in commentary, shared with us, shared openly, and shared in a way that other people could reference. It left a trail of thinking on that initial draft of the book.
We took all of that information back along with a more formal and more conventional, anonymous peer review. We had two processes going side by side. We took all that information, we sifted through it, we synthesized it, and we revised our writing. It was of course improved and strengthened, the crux of our book was really helpfully moved along the way, because of that social annotation process. It’s part of our work as well. It’s part of our workflow, if we’re going to talk about it, we need to do it ourselves.
Ross: One distinction is between accepting text and highlighting; for example, highlighting either with a highlighter in a physical book or by copying a block of noting in a Kindle or putting a text, I think that’s akin on the large level to social bookmarking, where you might just say, here’s an article, I’ll just make a note of that whereas the annotation is when you’re actually saying something about it: I agree, I disagree or to make some more points around that? Just to dig into that for somebody who’s trying to think about how they access information and build that into knowledge, that distinction between either highlighting, we’d love to know what language you might use around that, highlighting versus active notes which add to what is there.
Remi: Highlighting is a really important entry point into a richer repertoire of annotation practices. Highlighting is actually, I wouldn’t say controversial, but there’s mixed evidence; I’m a researcher, day to day, I’m a scholar, I read a lot of peer-reviewed research, and it’s actually pretty mixed evidence about the cognitive benefits of highlighting, I won’t get into all of the studies, but there are some recent ones, looking at undergraduate students, for example, who highlight course texts, then take quizzes, and their subsequent academic achievement, the evidence is rather mixed about whether just highlighting alone is a particularly effective cognitive strategy. However, in my world, it is one of a repertoire of annotation practices.
Of course, we can highlight texts but then we can also take those highlights and translate them into other kinds of notes, other kinds of annotations, including those that spark dialogue, or that allow us to confirm maybe that a particular interpretation is more or less accurate, or that we might be highlighting an instance of bias or disagreement, or we have a wondering, or maybe we’re highlighting because there’s a disciplinary method, maybe in the sciences, maybe there’s a particular type of rhetorical argument in learning about how to analyze essays that we’re curious about. Whether we’re in a course, we’re a learner, and we’re interacting with our peers, or we’re reading something else, maybe with colleagues, and we’re just doing something like a book club, but it’s digital, and it can be about sharing our notes. The highlighting is a starting point but it can lead to a much richer dialogue. That’s where the knowledge construction and the deeper synthesis of ideas begin to happen.
Ross: Let’s pull back to a user who’s never heard of annotation, they know what highlighting is, but framing this around, I call it knowledge development, you call it knowledge construction, the idea is we’ve got lots of information, information is just information, so this idea, how do we actively build that into our knowledge? Are there any practices or tools or starting points that you would suggest for somebody to do a more active annotation with the aim of knowledge development?
Remi: In this respect, what a really good place to start is with your own self-curation, which is that we like to write a lot; I think that one of the critiques of annotation is that it’s too messy, that there’s too much of it, that it’s just scribbling or it’s defacing a book, or it’s inherently a transgressive act because we shouldn’t write on the things that we read. I understand that, I can appreciate that, but when I say self-curation, it’s important to speak back to a text, and maybe just underline a few notes here, maybe it’s important to say ha, or I disagree, or I wonder, or leave little question marks, or leave little arrows, or smiley faces, almost like emoji, those kinds of symbols have been appearing in texts for hundreds of years. Having said that, if we look back at a corpus of annotation, we’ll find that there are maybe a handful of rough draft thoughts that can become then much deeper, more developed ways of thinking about a text. This is where the meaning-making process occurs.
If I spend an hour reading a text, and I’ve written all over it, many notes in many different kinds of ways, if I return to that text, at another point in time, I might find that there are just three or four that continue to speak to me, that continue to resonate with me, and I can then curate those, maybe I move them into another document, maybe they inform subsequent writing, maybe those are the types of notes that get shared socially, as opposed to that remain private. There are all kinds of ways in which as a first step, whether we’re writing by hand or writing on a computer digitally, we can return to our note-taking and annotation practices, and select those that remain resonant, select those that remain powerful to our thinking, and then inform subsequent work, inform the subsequent kind of writing, thinking, deeper production of knowledge, it is important for whatever work that we’re doing. That’s the first step: curate what you’ve annotated.
Ross: One frame around this is the same, annotation happens on a particular document, be that a book or a blog post, or whatever it may be. But these may be in the context of developing knowledge in a particular domain where you might want to learn more about diabetes or around the edge of artificial intelligence, all sorts of topics, in which case, you have a whole set of different texts, which you’re annotating. How do you then pull these together to build understanding around the topic where there are multiple texts?
Remi: This is a good point, Ross, because now we’re talking about annotation not just being tethered to an individual text, we’re talking about annotation being the connective tissue, or what some people refer to as the associative trails across multiple texts. Now, if we imagine a library full of books, we imagine our notes inside those books on shelves, it’s hard to imagine how those all get connected. Again, for hundreds of years thinkers and scholars have tried to imagine ways in which those notes become associated with one another and how across a domain, we can bring that thinking together in new ways. Of course, computational methods have helped with that incredibly, even in the last few decades now, we’re seeing great strides in that.
There’s a tool that I happen to use, it’s called Hypothesis. It’s an open-source, web annotation technology, it’s free, anyone can use it, and it can mark up and write on the entire web. That is one technical approach to finding a way in which wherever my notes may live, across whatever kinds of documents, blog posts, primary source of literature, news articles, e-reading platforms, open textbooks, individual correspondences, the historical record, if it’s digital, it’s living online, if it’s on the web, I’m now not just annotating a single document, I’m using an annotation platform that allows me to bring all of my thinking across all those documents and all of my annotations together in one place.
Again, I can choose to make those private just for me, I can choose to share those with other people. But it turns annotations from an isolated marginal comment into the connective tissue, those associative trails across all of my reading, across all of my thinking, and that allows me to then dig much deeper, as you said, particularly into domain-specific topics that require that kind of synthesis.
Ross: I want to just briefly diverge and say, that sounds very much like the concept of the global brain where all of the thinking of individuals coalesces into a higher order thought structure.
Remi: It does resonate. This is where again, I think that we find collaborative reading and writing technologies to provide glimpses of that. We could say that in this respect platforms even like Wikipedia, where you have multiple individuals reading together, editing together, and curating knowledge in rather sophisticated ways, although the platform is certainly not perfect, it has some documented issues. Nonetheless, it is a record, a visual record of people reading and then writing together. I think that we see other examples of that, and annotation and annotation platforms can serve as one type of that where the collective thinking of either a crowd, but also, in some cases, very highly trained expert communities can make new ways of thinking visible, accessible, and then really actionable to a much wider audience. Annotation is the key that unlocks that opportunity.
Ross: In the West, we have fairly recently seen Zettelkasten become very trendy, we’ve had Roam Research become a RoamCult, Obsidian, Logseq, and other tools, these all fall into what I call connected notetaking. This is perhaps a JSON, it’s not a sort of SQL thing but these are the tools that many people are using to be able to coalesce what they find, and what connections they see between them. How do these kinds of tools relate to what you’re describing in terms of annotation and social annotation?
Remi: In some cases, there’s quite a bit of overlap. I’m also quite interested in those note-taking histories because you mentioned Zettelkasten and many other methods around note-taking. In some cases, people are moving their notes off of a primary source. This is where again, it might not be appropriate or possible even to write on the text itself. They need to have a commonplace book or some other individual note-taking device, be it again, written by hand or online in some form, Roam being a great example of that, where notes can be curated, sifted through, where they can be used to be essentially remixed into other forms of writing, and they can aid in productivity workflows. I appreciate that. I understand that there is great value in that often for individuals, and often for individuals who are pursuing a particular line of inquiry.
I have my own bespoke note-taking practices where I like to jot my notes. When I’m talking about annotation, both in my work as a researcher, but also in my teaching… remember, I’m a university professor, I teach classes, I have students, I know that people are reading together so there’s a little bit of a shift in purpose here, between you, Ross, picking up a book or finding even a primary source literature or even a peer-reviewed article individually, and you reading about a particular topic that you’re fascinated with, macroeconomics, developments in political theory, whatever it happens to be, and you’re then taking your notes, wherever and however you do that.
The difference in purpose that I often find in my work is, let’s say, I have a group of 20 students, these are master’s or doctoral level students, and we’re going to read an article that explains a particular methodology germane to our field, we’re going to read it together and we have some shared learning objectives but we also know that our note-taking might move us in certain individual directions. But we have a common starting point and we’re also reading together, in that context, for that purpose; that’s where social annotation for me is particularly useful because I want that thinking to be made visible to the group. I want people to be reading together and then responding to one another, and I want us to be collectively making sense of that primary source.
There is then again, some overlap with some of the work that people are doing around common places like Zettelkasten, individual note-taking like Roam, etc. again, often for individual productivity. Much of the work that I do, which is we’re a group, either we’re a book club, or we’re a formal course at a university, or we’re a group of educators participating in professional learning, who are collectively working to make sense of common resources, common texts, to move our thinking forward, even if some of that is individually oriented, the group activity is part of what brings us all together.
Ross: And what tools do you use for that?
Remi: Again, Hypothesis has become a particularly useful tool because it’s free, because it’s open, because it works across a variety of education-specific platforms, but it happens to just live online on the web. There are also several other promising social annotation platforms out there that are bringing people together. I spend a lot of time working with primary and secondary educators, and K–12 education.
There is a nice annotation tool called NowComment that has become particularly useful. You’ve mentioned social bookmarking tools as well. Diigo, of course, has a long rich history. Others have been particularly generative in the formal education space. But this is where I try not to be too prescriptive about what tools people pick up because at the end of the day, it’s about social practices, how you want to be reading, thinking, writing, and collectively making sense of this stuff with other people? If you want to do it with other people, let’s figure out what those practices are, and then choose tools that meet our needs.
Ross: Do you use or have any thoughts around visual tools? For example, there are various plugins for Roam and Obsidian which give network representations of connections between ideas. TheBrain arguably uses… essentially mindmap collections. There’s also a concept mapping to be able to show concepts and the relationships between them. Do you use any visual thought mapping tools?
Remi: I do. I have a few of my own colleagues with several folks who have developed some of these tools as well to do this higher level thinking where you now are starting to establish associations between notes and resources, create broader categories of information, think about common themes, I think that that approach, particularly in visual ways is extremely helpful. The starting points are those initial rough draft thinking, that initial rough draft set of annotations. First of all, we need to create all this stuff. Again, this goes back to my earlier comments about curating even one’s own reading and one’s own initial annotations, it’s very important to have a corpus of rough draft thoughts, to begin with. Then whether those are mine, or yours, or a group of people’s notes, those can then all get put into other kinds of programs, other kinds of visual representational tools.
There are a few that I happen to appreciate and use, where you can literally without getting too technical, essentially export digital content into other applications. Then you can begin to remix those and say, which of these notes connect to one another? Which of these notes are useful to pair in a group? How does this particular set of comments help us to think more deeply about this particular question, this particular problem? I’m a huge fan of that. I think that there are some more, in this case, visually oriented, higher order thinking tools that can be used in this way, whether individuals are doing it, or groups are doing it. Again, the question becomes, where do you start with that process? You probably start by adding notes to texts so that you can even do that work in the first place.
Ross: What tools do you use or do you think are worth looking at?
Remi: There’s one, in particular, that is rather more academically oriented, its acronym is FROG. It’s coming out of a group of researchers that are combining the open annotation affordances of Hypothesis with some other visual platforms. I have a colleague Bodong Chen of the University of Minnesota, he is now off to Penn, it’s his research team and colleagues who’ve put this together. It’s pretty education specific, you need to have a few technical chops, but if you’re using, in this particular case, Hypothesis, to do that openly networked social annotation, you can then move your annotations into this separate application and then really participate in some very interesting work that helps to extend that beyond a collection of texts, still working at a group level to make that higher order thinking possible as you move into other knowledge construction activities, like essay writing or textual analysis.
Ross: This all goes to as I described synthesis is a human superpower, what distinguishes humans perhaps more than anything else, the idea, the ability to pull together disparate ideas into something which is cohesive, makes sense, and enables us to act effectively in a complex world. In addition to anything that we’ve already discussed, or perhaps to pull some of these things together, what are ways that you or you think that we can assist ourselves in that endeavor of being better synthesizers?
Remi: I think that’s a really powerful question, Ross. First of all, thanks for asking. I think of two things when I think of the ability to be better synthesizers. What is nuance? I think it’s really important, particularly in today’s information environment, whether we’re talking about the challenges of accurate information, conspiracy theory, or in general trying to find the signal through the noise, as I’m sure you’re aware of that catchphrase. It’s really important to act with nuance, to read with nuance, and to think about details specifically. Again, this is where annotation allows us to identify particular points of a text and particular aspects of an argument, even to the level of a character, certainly a word. We can say, here’s where I see evidence of this, here’s where I understand this argument building, here’s the year, the citation, we need to cite our sources, and we need to have very careful citational practices. That level of nuance can be aided in our work of annotation.
To move into synthesis, we can’t start from a level that is already too abstract. We can’t just try and synthesize big ideas and have more important insight from that if we’re not attentive to all of those small building blocks, all the little bits of nuance, all of those bytes of information that provide the strong foundation for our work. If you don’t have that strong foundation, if you don’t have that nuance, we’re not going to move to a point of synthesis. That’s the first big point I want to make. For me, again, practices of annotation allow me to be very nuanced in going back to information, finding particular aspects of a quote, or an argument of evidence.
I think the second big point I want to make, though, in terms of working towards synthesis is being able to then take a step back, and have a tool and a practice that allows us to make associations. Once we have all that nuance, once we have all the evidence, once we have our corpus of information, we need a good point of connection, we need an easy way, an intuitive way, a way that we can easily say, A connects to B, and A then connects also to C, maybe there’s a tenuous connection here, we need to be able to make those networks and we need to be able to make them visible for ourselves.
If we don’t have a connective practice, if we don’t habitually have some way of drawing associations, whether it’s visually, or it’s using a text-based platform, if that’s not a habit, and again, it’s going to be very hard for us to move into patterns of synthesis and to make those broader insights about the work that we do.
Connection making, although people say that humans are pattern recognition machines, if we don’t practice that, it gets pretty rusty. It doesn’t allow us to then make more sophisticated connections over time. We also need to make those connections habitual. That means that we need tools and we need practices. Without getting too jargony here, this is where academic literature tells us, shows us, there’s strong evidence of this, my work speaks to this, that knowledge construction practices can be learned, they can be identified, they can be built upon whether that means identifying points of bias, working through disagreement, building consensus, or asking questions, those are connections that we can habitually learn to make. If we do that over time, we can work towards deeper levels of synthesis, we can create those deeper insights in our work. Just to quickly summarize, we need to have that nuance and we also need to have those connections. Those are both habitual practices that we can as thinkers, readers, and writers develop over time.
Ross: If there are just a few of those references that you’d like to share with us, we’ll put them on the show notes because I certainly want to dig deeper, and I’m sure that some of my listeners will as well. Those are not something you can necessarily cover in a quick conversation.
Remi: Sure. One can’t do it, of course.
Ross: To round out, the topic is thriving on overload, which is something I think you do and you assist your students to do. Could you summarize just a few top-of-mind recommendations, to-dos, or ways in which people awash with information can pull that together to be on top of that, to feel that they are prospering in that way?
Remi: Read slowly. The first thing I’ve done in my courses, particularly with my students over the past few years, is to eliminate the volume of what we read. I want us to read more deeply, I want us to read more deeply and I want us to read fewer, whether it’s books, texts, articles, whatever it is, just really winnowing down, first of all, the amount of information that we read so that we can read more slowly. Then as we read, again, it’s nice to read together. When we read alone, there is a lot to be gained and it can be very personally meaningful but the benefits of reading socially are incredible.
The benefits of reading together with other people are to check bias, remind us of perspectives that we might not have considered, to think more deeply about questions that we might not have asked, which happens when we read socially. That’s the second opportunity. How do we thrive on information overload? We can read with other people because they can keep us in check and remind us of why we’re reading perhaps this in the first place. Then a third recommendation as we’ve been talking about is having robust note-taking practices. Again, if that’s going to be more individually oriented, with a commonplace book, Zettelkasten, whatever it may be, or you’re taking notes with other people as a form of discussion and conversation, having a robust annotation practice can be a really helpful way of thriving on information overload. But reduce the volume, reduce other people, and have a robust note-taking and annotation practice, that’s a nice recipe for thriving on information overload.
Ross: That’s fantastic. I think that’s a very distinctive set of recommendations that are extraordinarily valuable. Thank you so much for your time and your insights, Remi. It’s been a really fascinating conversation.
Remi: Ross, you are so welcome. I hope that listeners find this helpful. There’s so much information out there. I’ll share some resources and people can follow my work, but I appreciate the opportunity to chat. Thank you so much.
Ross: Thanks.
The post Remi Kalir on social annotation, self curation, the connective tissue of ideas, annotation tools, and nuance for synthesis (Ep31) appeared first on Humans + AI.