Argument Mining

In 1987, computer scientist Robin Cohen outlined a theory of argument structure which laid the groundwork for modern argument mining tasks. Taking argument to a process in which a speaker intentionally tries to convince a hearer, her approach focused on understanding the structure arguments can take.

This structure is generally tree-like: the speakers primary claim is the root, and supporting arguments appear as branches. Secondary arguments may further expand the tree, as the speaker makes claims to reinforce a supporting argument. That is, a simple argument can take the form A and B, therefore C, or could take the form A therefore B therefore C.

In this way a complex argument can be modeled a tree with all the various supporting and secondary arguments point back up to the core argument root.

The problem that Cohen noted, which has continued to be a challenge in more recent argument mining techniques, is that core premises often go unsaid.

Take, for example, the simple argument structure of “P therefore Q.” In many contexts, a speaker will state P and Q, but leave out the primary claim: P therefore Q. As human interpreters, filling this gap is often a trivial task. Consider the simple argument:

Joey is dangerous.
Joey is a shark.

It is left the reader to infer that Joey is dangerous because he is a shark…and that all sharks are dangerous. (This, of course, could be debated…)

While there are no doubt instances where this lack of clarity causes confusion for a human reader, in general, this is a challenge which is easy for people with their broad array of contextual knowledge – and terribly difficult for machines.

Joel Katzav and Chris Reed formalize this missing argument (enthymeme) challenge. Defining an argument as “a representation of a fact as conveying some other fact,” a complete argument then has three elements: a conveying fact, the appropriate relation of conveyance, and the conveyed fact.

In parsing content, then, an algorithm could work to define a sentence or otherwise defined element as either a “non-argument” or as one of the argument types above. This makes the computer’s job a little easier: it only has to recognizes pieces of an argument and can flag which arguments are incomplete.

Furthermore, syntactic clues often give both humans and machines some insight into the structure of an implied argument: because X, therefore Y. Annotated debate texts can then help machines learn the relevant syntactic clues, allowing them to better parse arguments.

This is still somewhat unsatisfying, though, as annotating texts is difficult, expensive…and may still be inaccurate. In one study of online-debate, Rob Abbott et al employed 5-7 annotators per post and still found not-insignificant disagreement on some measures. Most notably, it seems, people are not much better at recognizing sarcasm than people.

Furthermore, arguments are not always…formal.

In legal texts or a public debate, it might be reasonable to assume that a given speaker makes the best possible argument as clearly as possible for a general human audience. This assumption can not be extended to many online forums or other domains, such as student essays. Colloquially, syntactic clues may be missing…or may even be miss used.

Latest work in argument mining has focused on over coming these challenges.

A 2015 paper by Ivan Habernal and Iryna Gurevich, for example, aimed to build an argument mining system that could work across domains, on unlabeled data. An earlier paper by Christian Stab and Iryna Gurevich focused on trying to parse (often poorly-formated) student essays.

By projecting argument elements into a vector space – or argument space – researchers can use unsupervised techniques to cluster arguments and identify argument centroids, which represent “prototypical arguments” not actually observed in the text.

There’s still more work to do, but these recent approaches have been reasonably successful and show a lot of promise.

facebooktwittergoogle_plusredditlinkedintumblrmail

Using Ground Rules to Create “Safe-Enough” Spaces

We learned a lot from the article below written by a team from the Public Conversations Project, one of our NCDD member organizations. The piece uses story and art to offer a valuable lessons about how ground rules in dialogues can temper the impacts that power and privilege frequently have on tense discussions and help everyone be heard, even when emotions run high. We encourage you to read the PCP article, cross-posted below, or find the original here.


No, We Won’t Calm Down: Emotion and Reason in Dialogue?

PCP new logoA recent cartoon on digital platform Everyday Feminism stimulated a lot of questions among Public Conversations Project staff. Entitled “No, We Won’t Calm Down-Tone Policing is Just Another Way to Protect Privilege,” it raised important issues about power, privilege, the apparent contrast between reason and emotion, and the roles of advocacy and dialogue.

Tone policing

The protagonist, Robot Hugs, talks about how tone policing allows privileged people to define the terms of a conversation about oppression and how this “hinges on the idea that emotion and reason cannot coexist – that reasonable discussions cannot involve emotions.” It further asserts that this allows privileged people to regain control of a conversation that is making them uncomfortable and thereby avoid the discomfort caused by being exposed to the very real emotional fallout of oppression and discrimination.”

Image via EverydayFeminism.com, Credit: Robot Hugs

Our dialogue work frequently focuses on polarized and extremely controversial topics that touch on issues of power and privilege. We see “tone policing” as something to be avoided; we value people’s bringing their feelings into dialogue. That is one reason we talk with participants beforehand: to offer guidance about how they can speak in ways that are more likely to be heard, and how to listen with resilience. The communication agreements that participants commit to beforehand are ones that they have jointly drafted and found acceptable to support their purpose in having a deeper, more authentic conversation.

Power and privilege

The cartoon raises challenging and important questions about power and privilege that surface frequently in the course of our work. In fact, many partisans (on whatever side of a controversial issue) see advocacy and dialogue as mutually exclusive and cite issues of privilege and power imbalance as reasons it should be avoided. As Robot Hugs continues, “these conversations aren’t meant to be comfortable. We are discussing real, dangerous, structural things that make lives worse for entire groups of people. If it makes you feel uncomfortable, the thing to do isn’t to try to get us to talk about it differently – the thing to do is to help us stop it from happening.”

Why there may still be a need for dialogue

There are, however, many occasions when people on different sides of important issues feel the need to sit down and talk together. They may be tired of conflict or violence, or may see the potential benefit to their community of such a conversation. Our work in Montana was initiated because pro and anti-open carry advocates decided that it was important to try to understand each other’s perspectives, to strengthen their communities and keep them safe. In Nigeria, we worked with Muslims and Christians who wanted to address sporadic outbreaks of violence between their communities.

Bottom line, there are times when people experience the need to listen to one another, as a first step toward building relationships and trust. If such efforts succeed, people may be interested in attempting to work together to address problems, when neither group can solve the problem on their own.

Shared purpose, shared power

A dialogue is a conversation that participants enter with the clear and shared purpose of mutual understanding. They have also had an opportunity to contribute to how they want to be together. They contribute ideas that will promote this purpose, so they are the ones who are actively participating in designing the structure and communication agreements, rather than someone who is more privileged or powerful “imposing” them. They have agreed to focus on certain questions, to limit the time of responding, and to respond in ways that enhance learning and connection. We are aware that everyone is not always interested in such conversations and they may not be possible in certain circumstances.

The power of agreements

The agreements are co-created by all participants so that those who are “privileged” hold the same power as all others. Agreements can be negotiated throughout the process, so that if something is not working, the opportunity to fix it exists. In dialogue, the purpose of the conversation is mutual understanding. We know that when people are having difficult conversations around polarizing issues, it is helpful to create a space for effective communication. Hard work! In order to create a safe space – specifically one that does not induce a flight, fight, or freeze response – a person has to feel safe.

Avoiding fight, flight, or freeze

A part of our brain is watching for danger, and may prevent us from being capable of having a constructive conversation when we most need it. When there’s a lot at stake and we feel under attack, the brain and central nervous system release hormones designed to keep us hyper vigilant, with physiological (racing heartrate, cold, sweaty palms, etc.) and psychological effects. Our capacity to think and reflect shuts down as we prepare for fight, flight or freeze. A conversation with highly emotional responses, however justified, can trigger this reactive response. A structured, voluntary conversation, however, creates a sense of safety and wellbeing so that participants can focus on the narratives and not the fear that emerges because of feeling threatened.

One example of this dynamic occurred a number of years ago, in which issues of power imbalances and privilege caused collaborative work to run aground and precipitated a request for our assistance. The Massachusetts Department of Mental Health had received a three-year federal grant to reduce the use of seclusion and restraint in its psychiatric hospitals. DMH formed a Steering Committee, which included people with lived experience of psychiatric illness, family members, advocates, mental health clinicians, hospital directors, and staff.

Although all the participants shared a common purpose, the enterprise foundered within its first six months, as Steering Committee members experienced massive frustrations, with many voicing the sense of not feeling seen or heard by others. People with lived experience spoke powerfully of their sense that DMH staff were unwilling to hear their experience of having been traumatized by the seclusion and restraint orders that psychiatric hospital staff had initiated. In response to voicing their concerns, the message they heard back was that they needed to speak differently so that the (more powerful) DMH staff would not feel attacked. Many of the clinicians felt guilty and misunderstood, seen as one-dimensional and complained of being attacked verbally when they attempted to engage or empathize. From their perspective, there was a power imbalance in terms of the “moral power of the victim.”

Freedom through structure

One of our first tasks was to help them figure out how both sides could express themselves clearly and powerfully, in ways that invited thoughtful listening, rather than resistance and shutting down. How could the issues of power and privilege be addressed in a way that would allow them to resume their work together? We began by meeting with each group separately and helping them think through their priorities and their purpose in coming together. It soon became evident that there were significant differences within each group.

We encouraged a candid discussion that focused on helping them identify the kinds of behaviors and commitments that would support their purpose. As participants explored their own feelings and experiences within each group, they engaged energetically with each other. As facilitators, we did not impose “ground rules” but allowed these to emerge from the group after thorough discussion.

When the two groups came together one of the first items of conversation was the negotiation of these ground rules. This was accomplished quickly and it successfully provided the kind of “safe-enough” space within which participants were able to have a more fruitful conversation that led to their getting back on track.

Purpose first; no policing later

Image via EverydayFeminism.com, Credit: Robot Hugs

To return to Robot Hugs, one of the underlying assumptions that we noted was the lack of clarity in identifying the purpose for the conversation. Robot Hugs expressed the belief that sometimes conversations are not just for moving toward solutions but they can also be for exploring situations, letting off steam, finding community, and feeling less alone. The cartoon suggested that those conducting the “tone policing” had very different purposes for the conversation, namely to retain their power and privilege and avoid feeling uncomfortable.

One thing that we emphasize in our work is the importance of purpose. If the purpose for a conversation is clear and shared, then developing shared commitments of how people want to be together can help support that purpose. If the participants have very different purposes for the conversation, these differing purposes frequently manifest behaviorally and interfere with task accomplishment. If the purpose involves mutual learning and understanding, differences of power and privilege can usually be directly addressed and successfully negotiated. This kind of direct and open conversation, focused on mutual learning and understanding, can lay the foundation for collaborative action to create more fairness and justice.

Robot Hugs rightly complains that the more powerful people sometimes want to make the rules of the game, to impose their views on how the conversation should be held. In our work, communicating with participants beforehand helps to address issues of purpose, conducive behaviors, and commitments. Groups might, for example, talk about sharing airtime, listening with resilience, and other kinds of behaviors that would help support their purpose.

Emotion and reason are not enemies. We want people to bring their feelings and their passions into the room when they engage in dialogue, but not to be overwhelmed by them. And we want to help them to think about how they can express these in ways that invite listening with an open heart and speaking in ways that invite receptivity.

You can find the original version of this Public Conversations Project post at www.publicconversations.org/blog/no-we-wont-calm-down-emotion-and-reason-dialogue#sthash.YJKzIe1V.dpuf.

Learning

I wonder if the process of learning is like…sediment on a shore.

That doesn’t sound very glamorous, but it feels appropriate somehow.  A wave comes in, carrying all sorts of knowledge – far more than one person could possibly manage. It’s a little overwhelming. You might lose your footing. Or recklessly risk being swept out to sea.

It’s exhilarating.

And then the wave recedes, eclectic flotsam left in its wake.

You gather up what bits you can; painfully little compared to the vast sea before you. And you wait for the next wave to come in; awash with possibilities.

A good class is like a good book: once you finish it, you want to read it again; to rediscover its mysteries anew.

facebooktwittergoogle_plusredditlinkedintumblrmail

Alabama Prisons: Why We Cannot Look Away from Alabama’s Shame (DMC Issue Guide)

The issue guide, Alabama Prisons: Why We Cannot Look Away from Alabama’s Shame, was a collaborative effort between David Mathews Center and AL.com, published 2014. The guide offers three approaches for deliberation to address the serious and widespread issues with the Alabama prison system. In addition to the guide, an eight and a half minute video was also created to summarize the realities of the Alabama prisons.

The guide offers three approaches for deliberation and within these approaches are five specific actions and consequences for each option. Below are the three approaches from the issue guide which were found on National Issues Forums Institute blog here. You can find more information about the issue guide, including the action/consequence of each approach and the brief video, on AL.com here.

From NIFI blog…

Approach One: “Increase Capacity and Improve Basic Conditions”
Alabama’s prison population far exceeds operational capacity, and conditions inside the facilities are raising constitutional questions. As a result, the Alabama Department of Corrections is now facing potential federal intervention and costly lawsuits. If the state does not significantly reduce overcrowding and improve basic conditions for inmates, then additional lawsuits may be filed and thousands of prisoners may be released. People want to feel safe in their communities, and many residents and lawmakers want to ensure that our prison system complies with the Constitution. If we want to avoid lawsuits, federal intervention and a potential release of prisoners, then we must increase capacity and improve conditions in the prisons. Our time line is limited and addressing difficult sentencing issues and root causes may take too long. If we want to solve this problem and stay tough on crime, then we must consider building new prisons, expanding existing facilities, and/or contracting with for-profit prisons.

Approach Two: “Address Root Causes through Education, Support and Rehabilitation”
Thousands of Alabamians are incarcerated every year, and the prison system is under stress. If we want to truly address the overcrowding issue, we cannot simply build more prisons. People must work to understand the root causes that lead residents to commit and re-commit crimes, and provide support to help remedy those deeper issues. Many people need educational support, community-based mentoring, substance abuse counseling and mental health services. Many offenders need access to educational services, job training and behavioral health support while incarcerated and after release. If we want to keep people out of prisons and avoid high recidivism rates, then we cannot ignore the real issues that drive individuals to break the law. By providing education, support and rehabilitation, we may also see benefits to communities, families, and the economy.

Approach Three: “Implement Alternative Approaches to Incarceration”
People who break the law must face consequences. Unfortunately, many Alabamians who break the law end up in state prison — resulting in overcrowding and dire conditions. Alabama’s increasing prison population is costing taxpayers a significant amount of money, and the long-term impact on communities and families is troubling. If we want to continue to punish criminal behavior and avoid the costly practice of mass incarceration, then we must consider alternatives to prison. Specifically, we must implement community corrections programs in every county, expand problem-solving courts and provide opportunities for restorative justice. We must also ensure that justice is applied in an equitable and consistent manner.

About DMC and the Issue Guides
The David Mathews Center—a non-profit, non-partisan organization—authors deliberative frameworks for people to carefully examine multiple approaches, weigh costs and consequences, and work through tensions and tradeoffs among different courses of action to current and historic issues of public concern.

David Mathews Center issue guides are named and framed by Alabamians for Alabama Issues Forums (AIF) during a biennial “Citizens’ Congress” and follow-up workshops. Alabama Issues Forums is a David Mathews Center signature program designed to bring Alabamians together to deliberate and take community action on an issue of public concern. Digital copies of all AIF issue guides, and accompanying post-forum questionnaires, are available for free download at http://mathewscenter.org/resources.

Follow DMC on Twitter: @DMCforCivicLife

Resource Link: www.nifi.org/en/groups/online-issue-framework-about-alabama-prison-reform

Policy Forum

Method: Policy Forum

The following is a stub, please help us fill it out. For an example of a policy forum see: http://everyday-democracy.org/resources/strong-starts-policy-forum-repor... Definition Problems and Purpose History Participant Selection Deliberation, Decisions, and Public Interaction Influence, Outcomes, and Effects Analysis and Lessons Learned Secondary Sources External Links Notes

Dr. Randy Felton: A Celebration of Life

Something personal on the blog here.

Dr. Randy Felton was a friend and somewhat of a mentor to me. He just recently retired as the Test Development Center’s social studies coordinator. This was basically a job he took to do a favor for the state of Florida and his colleagues in the social studies. He was an excellent state level social studies specialist, a classroom teacher, a district leader, and a visionary. I first met Randy when I worked on developing new world history standards almost ten years ago now. I had the pleasure of working with him on Florida’s US History assessment, writing items for Pearson and working with Randy through training and collaboration. I most recently had the opportunity to spend time with him during Civics EOC item review back in October, a fun but intense week of work. He was an incredible leader in the social studies here in Florida and such a pleasure and joy to work with, to know, and to yes, drink beer with (something we definitely had in common). He was a Seminole and a Navy guy, and I am a Gator and an Air Force vet, and yet we got along great and I learned a great deal from working with him. His loss will be felt. Deepest sympathy to his family and friends, both from me personally and from Doug, Peggy, Val, Terri, Elizabeth and everyone at the Florida Joint Center for Citizenship.

This link provides access to 200+ fabulous photos from an amazing life.


Register for the Intro to NCDD Webinar, THIS Wednesday!

As we’ve previously announced, we are excited to be hosting our first webinar introduction to NCDD this Wednesday, April 20th from 1-2pm Eastern / 10-11am Pacific, and we encourage you to register today to join us! This “get to know us” event is a great way for D&D newcomers and veterans alike to learn things about NCDD’s Small green NCDD logowebsite, resources, network, and staff that you might have never known. You won’t want to miss it!

During the webinar, our five NCDD staff members will take participants on a virtual tour of the NCDD website including the blog, resource center, events pages, the member directory and map, and the listservs. We’ll also include a run down of how to best use our social media resources, as well as some little-known tips and tricks for getting the word out about your work.

This intro to NCDD webinar is the perfect chance to make sure you know how to take advantage of all of the tools your NCDD membership offers for learning about the D&D world and connecting to others in the field. Plus you’ll get to meet the wonderful people on the NCDD team who help the whole thing run. But you time is running out, so you have to register today!

We especially encourage new members and folks interested in joining NCDD to attend, but the call is open to everyone! It will be a good refresher even for the most experienced members. on how to get the most out of NCDD’s website. Be sure to register today!

Clarence Thomas’s Black Nationalist Jurisprudence

I don’t know a lot about the role of Anita Hill in Clarence Thomas’s Supreme Court confirmation hearings. I was just a bit too young to understand what was happening, and so I’m looking forward to watching the new HBO movie on that topic. Thomas himself famously called it a “high tech lynching.” My suspicion is that Thomas probably was guilty of sexual harassment, but that there was almost certainly a concerted effort to link it to his race in ways that we should find abhorrent. On the other side of the aisle, Bill Clinton certainly seems to have been guilty of similar activities.

In the spirit of preparing for that viewing, I’m revisiting an old post of mine on Clarence Thomas’s Counterrevolution wherein Corey Robin (of The Reactionary Mind fame) discusses the intellectual legacy of Justice Clarence Thomas:

“The first time Clarence Thomas went to Washington, DC, it was to protest the Vietnam War. The last time that Clarence Thomas attended a protest, as far as I can tell, it was to free Bobby Seale and Erikah Huggins.”

Reading Robin’s efforts to make Thomas intelligible always has me worrying about why African-Americans don’t take anti-statist positions more often (libertarianism, anarchism, localism.) After all, you can think that there is a government harm without thinking there is a government solution.

That’s basically the heart of the Black Panthers’ ideology: the white police can terrorize Black communities, but they can’t fix them. On the Black Panther view, only Black people can fix their own communities, because when white people try to do it it’s just as cover for more colonialism. Which is why their most powerful and dangerous program wasn’t the guns, it was the Free Breakfast for School Children program.

Tally up the affirmative action and racialized mass incarceration; the public housing and the brutal policing; the food stamps and the white social workers; the Black Presidency and the decades of shame and racism in every government office; integrated schools and two-hour bus rides that made accountability impossible; racial gerrymandering and the destruction of most of the neighborhood non-religious institutions in African-American communities.

Now ask yourself: did African-Americans come out ahead?

Becky Pettit did the math, and she says no. Lots of African-Americans are skeptical that a white supremacist government will ever offer solutions, simple or complex, that are actually tailored to the needs of their community. Perhaps the skeptics are wrong. Perhaps we can do better. But we haven’t so far. We’re again talking about a New Deal for whites, without ever having had a new deal for Blacks.

Many white conservatives (and more than a few white liberals) think these disparate outcomes can be explained because Blacks are inferior. A lot of movement conservatives follow either Charles Murray or Amy Wax on this: it’s either a natural inferiority (Murray) or a cultural one (Wax.) But Black conservatives like Thomas seem to think it’s something else: they think we white liberals just keep making the situation worse, and that we don’t care so long as we can assure ourselves that we’re well-intentioned.

And our treatment of Thomas is part of the problem: we ignore or disdain him, or worse, insinuate that he’s a race traitor who can’t recognize a simple self-contradiction. How could he come out against affirmative action given that he benefited from it? Consider the Parents v. Seattle concurrence and the Zelman concurrence. These are serious acts of jurisprudence: among other things, Justice Thomas points out that integration destroyed centers of Black excellence, took excellent Black teachers out of Black classrooms and put them in White classrooms, leaving majority Black classrooms with substandard instruction, frequently by whites who didn’t understand the culture and harbored unconscious racism.

On Thomas’s view, the 14th amendment isn’t race blind; it proscribes race consciousness except when responding to state discrimination. In Zelman, Thomas says that the history of poor performing minority schools DOES count as state discrimination in need of remedy; and in Seattle he says that the kinds of remedies that can be used should never include the individual racial classification of elementary school children. Lots of other remedies are still possible, but don’t go creating huge bureaucracies of racial classification, because that hasn’t tended to be a good thing in US history. Thomas’s point is that it’s not enough to say “this time it’s different,” because that’s what whites always say.

Has race-conscious hiring through affirmative action helped African-Americans? Certainly in some cases! And yet at every income level, Blacks have higher levels of educational attainment than whites, while they also have higher unemployment for their educational attainment than whites. Their unemployment rates reflect longer searches, with less success, even with equal or better qualifications. These are facts that should make us blush with shame, and panic that our worldviews and preferred remedies are inadequate.

It’s no better in universities, but here is where I begin to part ways with Thomas. We’ve gotten to such a weird place on the court’s race decisions that we’re forced to debate the benefit to white people of having African-Americans and other racial minorities in the classroom instead of remembering the serious racism that infects all of our institutions.

Cribbing from Elizabeth Anderson here, Fisher and the university-level affirmative action cases are a total muddle for advocating diversity for its own sake. The liberal justices have been backed into a corner and it seems that they know it. The real problem the court faces in cases like Fisher, Zelman, Seattle, etc. is justifying affirmative action or individual racial classification or school choice in the absence of de jure discrimination. Which is dumb, because we had a whole lot of that de jure discrimination not so long ago, and it clearly still echoes into the present.

The long history of de jure discrimination implicates every US institution, regardless of particulars, but the whole of the Supreme Court has disagreed with that reading. Thus, they’re having a pretty silly argument in cases like Fisher: instead of saying, “Hey, remember slavery, segregation, and the KKK? They’re still relevant here!” the court says, “Well, forget about slavery, segregation, and the KKK: wouldn’t it be nice if we had more skin colors and ethnic groups in Freshman composition?” The real problem is that everyone on the court accepts that “Forget about it!” rubric for thinking through racial remedies. That problem goes back to Milliken v. Bradley and the refusal to enforce interdistrict busing.

Yet Thomas is the one Justice who doesn’t forget about that history and continually brings it up. And what he says in cases like Seattle and Zelman is: as a Justice, it’s my job to keep my eye on the de jure. De jure justifications at the SCOTUS level affect national policy for centuries, and we can’t afford to make exceptions for racialized classifications just because the individuals involved think they’re doing the right thing. After all, that’s how we got segregation and internment camps in the first place.

And here’s where I think Thomas is offering a powerfully Black Nationalist perspective: he seems to me to be saying that as a matter of policy we should remember that we live in a white supremacist society, that Blacks are in the minority and the political institutions will usually be controlled by white politicians and white voters. So always be careful of the tools and remedies you make available to those white racists, because they’ll use the rubric of the white man’s burden to justify all sorts of evils. They’ll use well-meaning liberals as the tip of the spear, like colonial powers used missionaries or US neo-imperialism uses the Peace Corps.

I mean, as a white liberal, I find myself worried about the ways good intentions and reform get twisted into such terrible institutions for non-whites, as they’ve been twisted over and over and over again. We’re often told that when we tally up the costs and benefits of government we should ignore the effects of the police and prisons because they’re not justified in terms of helping African-Americans. But of course they are: law and order rhetoric is popularly justified by pointing to Black victims and support without noting that shock-and-awe policing and mass incarceration are a poor remedy that creates as many victims as it avenges.

Background