How can we assess with any accuracy whether someone is lying to us through online communication media? And - what can we do to prevent it? This article spotlights an issue that anybody practicing ODR needs to consider.
The concept of interparty trust has often been raised with regards to ODR. This topic has been a central discussion forum in Cyberweek for the past several years, and is the topic of many articles on ODR.
The perspective taken has usually that of building trust: How can I get the other party to trust me, online? How is trust-building different in online processes, compared to fa...? What might a mediator do in order to promote interparty trust?
Less attention has been paid to the flip side (perhaps, the Dark Side) of this topic: Setting aside our desire to engender trust in others, how do we know whether to place our own trust in them. Should we, indeed, trust others online? Whom should we trust? How can we recognize that most basic act of untrustworthy behavior: deception?
A recent article by Brian Farkas in the Cardozo Journal of Dispute Resolution entitled “Old problem, new me...” seeks to shed light on this topic. The author reviews recent research on deception and lying in face-to-face negotiation, and discusses its application to online processes. Consiering the frequent incidents of internet scams, hoaxes and frauds, coupled with the booming spread of online commercial negotiation, this is clearly a piece whose time has come.
Relying on an article by Peter Reilly exploring the nature of lying in negotiation and mediation, Farkas explains the differentiation between lying and deception: The former is actively making a false statement; the latter is any other method of concealing the truth – including acts of omission. One way or another, parties often lie to each other during ADR processes, and no normative actions are likely to rein this in. How might one go about minimizing the risk of their opposite party lying, or maximizing the chances for accurate recognition of deceptive moves?
Farkas suggests that the literature divides into two types of approaches: Relationship-based, and scientific.
Relationship-based: Peter Reilly prescribes two strategies: Conduct thorough background research before the negotiation, and engage in long-term relationship building with likely opposites. The former should include not only the substance of the deal, but also the counterparty’s reputation. The latter is likely to provide you with an emotional and behavioral ‘baseline’ for your counterpart, allowing you to recognize sudden changes in their tone or actions; such changes are likely to be indicators of deception.
Scientific: Farkas describes the work of two scholars, well-known for their work on emotions as well as on negotiation, Paul Ekman and Clark Freshman. According to these experts, while most of the pop-science body-language tips regarding lying are not scientifically evidenced, the emotional reactions of the deceiver do cause brief flashes of facial microexpressions which cannot be hidden.
These approaches – the relationship-based and the scientific – are the only ones that have any real chance of coping with deception. Regulatory and normative approaches to curbing negotiator deception have limited effect. The main players in this area are bar associations, whose codes of conduct bind only attorneys. Attorneys are often not the key players in commercial negotiation, and in these situations only vague and subjective “good-faith” clauses might be in place to constrain deceptive behavior.
Having explored deception and deception prevention in negotiation, Farkas goes on to explore how these issues play out online. While there is little research available, Farkas notes two significant research findings that can shed light on this corner of e-negotiation:
1) Humans are far less skilled in detecting deception online, than they are in face-to-face settings. In an experiment comparing deception detection through instant messaging as opposed to in in-person meetings, parties correctly identified their opposite as deceptive 9% of the time online, as opposed to 24% in-person (Giordano, 2011). If your counterpart lies to you online, there is very little chance you will detect this.
2) Linguistic analysis of text-based negotiations has shown that there are some traits associated with deception. Deceptive negotiators wrote more words, related to senses more often (e.g., seeing, touching), used fewer self-oriented words (“I”, “me”, etc. and more other-focused words (“you”) when lying than when telling the truth”. In general, they simply tended to be more long-winded (Hancock et al, 2008).
In addition, Farkas branches off into other challenges to successful online negotiation – the difficulty of building trust online, a tendency to increased positionality in online interactions, and the obstacles to mutual understanding based in language and culture which characterize so many online interactions. While these do not directly relate to the issue of deception, I think they support the main theme in the sense that they make it more difficult for negotiators to correctly identify online deception. Instead, we identify trustworthy moves as deception, and deceptive statements as misunderstandings or truthful communication.
I’ll point out that about ten years ago, early research on this topic (Thompson and Nadler, 2002) showed that indeed, online negotiators are more likely to suspect that their counterpart is deceiving them (compared to face-to-face negotiators) – even when no actual deception has taken place. In other words, our problem is not merely detecting lies, it the overall accuracy (or lack thereof) of discerning truth from lie in online interactions. Our low accuracy, coupled with the increased prevalence of lying by online negotiators, pose significant challenges to online interactions.
Farkas points out that the lack of physical interpersonal contact in ODR challenges the scientific approach to identifying deception – the ‘tells’ are all in the body, not in the text. (I’d suggest that as we become more skilled at online negotiation, we will be able to recognize textual cues such as those suggested by Hancock much as we are able to recognize a change in tone or a perspiring face). The relationship-based strategies are challenged as well, given that online negotiators tend to pass up on opportunities for relationship building, empathy, open-ended questions and the like.
What can be done to limit and/or recognize deception online, nonetheless? Farkas recommends making early and repeated reference to standards or objective criteria – in an attempt to limit the “playing field” of acceptable statements. Additionally, direct and to-the-point questions might assist in avoiding lies of omission, particularly in the textual medium where the question doesn’t dissolve into the air but remains on the record.
I’ve stated elsewhere that ODR is definitely still at a text-centered stage; however, use of video is likely to become more pervasive. Farkas points out that no research has addressed the efficacy of video-based communication for assessing deception, but suggests that high-quality video –based ODR – or even holographic interactions - might deliver the goods in this area.
I’d suggest that building on Farkas’ suggestions about the prevalence of deception in general, and the challenges to it online in particular, researchers would do well to zoom in on individual media and explore the rate of deception and the accuracy of deception-detection afforded by the media. Some work on this has been done – for example, experiments have shown that people lie more in emails than they do in pen-and-paper letter exchanges (Naquin, Kurtzberg & Belkin, 2010). Interestingly, in a recent TED talk, Jeff Hancock of Cornell University (who identified the linguistic footprints of text-based liars, as discussed above) suggested that while despite the bad rap online communication has received, people might actually lie less using certain text-based and social media. He demonstrates this in several specific media: People lie the most on the phone, far more than in email communication. LinkedIn resumes have been found to be more honest than traditional paper resumes. While people lie in online dating descriptions of themselves – they only do so by a little. Facebook personas and real-life personas correspond far more than many of us suspect. Why this tendency towards truthfulness? Hancock suggests that the searchability and permanence of information online – not only are the words written, but they are in the possession of the other and perhaps of third-party observers, keeps us honest.
Farkas ends with this caveat, which everybody studying deception would probably agree with:
“While ODR presents an exciting and efficient alternative for commercial actors, the platform does not change the simple truth that deception is always lurking. It is human nature. As Peter Steiner’s notorious New Yorker dogs remind us: Everybody lies.”