Expectancy violations theory holds that people arrive at an interaction already expecting a certain kind of behaviour, that a departure from the expectation is arousing enough to start an appraisal, and that the outcome of the appraisal turns on the direction of the departure and on how the person who departed is regarded. Judee Burgoon built it on nonverbal behaviour between people in a room, so a written conversation carries some of the behaviours it names and none of the expectation they departed from.<\/p>\n\n\n\n
The theory began with how close people stand. Burgoon and Stephen Jones synthesised the proxemics literature in 1976 and placed it in a framework based on norms and expectations, advancing a model for predicting the effects of violating proxemic expectations, in Human Communication Research<\/a>. Everything the theory became afterwards is an expansion of that single move: stop asking what a behaviour is, and ask what it departed from.<\/p>\n\n\n\n In 1993 Burgoon and Beth Le Poire induced their participants to hold positive or negative expectancies about a partner, covering both the partner’s general personal attributes and their specific communication behaviour, before any discussion happened. Only then did the participants interact with a confederate who communicated in a pleasant, involved fashion or its opposite. The report of that study<\/a> says the preinteraction expectancies, and the personal attribute ones especially, caused perceivers to evaluate the targets and their behaviour differently, and that a negatively valenced expectancy worked as a negative violation in its own right.<\/p>\n\n\n\n The design is the useful part. The expectancy was installed first and measured as a separate thing from the behaviour, because in this theory it is a separate thing. An expectancy is not an inference drawn from what someone just did. It is already in place when they do it.<\/p>\n\n\n\n What people actually expect had to be established by measurement rather than assumed. Burgoon and Joseph Walther wrote in 1990 that there was a shortage of empirically based knowledge about which interpersonal behaviours count as expected and how they are evaluated, and ran a two part experiment with 622 respondents who rated the appropriateness, typicality and desirability of seven forms of touch and nine combinations of posture and proximity, reported in Human Communication Research<\/a>. Some behaviours came out expected and positively valenced; others qualified as positive or negative violations.<\/p>\n\n\n\n Burgoon’s own 2015 summary of the theory, written for The International Encyclopedia of Interpersonal Communication<\/a>, puts the mechanism in one sentence: human interactions are strongly governed by expectations which, if violated, are arousing and trigger an appraisal process that may be moderated by the rewardingness of the violator, and the interpretation and evaluation of the violation determine whether it counts as positive or negative.<\/p>\n\n\n\n There are three steps there and they are easy to run together. A departure happens. The departure is arousing, which is what makes it register rather than pass. Then an appraisal decides what it meant. The arousal step does not know yet whether the news is good.<\/p>\n\n\n\n Researchers testing the theory had to show that step two had happened before they could test step three. In two experiments on conversational involvement reported in the Journal of Nonverbal Behavior<\/a>, Burgoon, Deborah Newton, Walther and James Baesler record that the involvement changes they introduced were sufficiently unexpected, arousing and distracting to qualify as violations of expectations. That check is not a formality. A departure nobody registers is not a violation.<\/p>\n\n\n\n The prediction that sounds wrong until you read it twice is this one, from the 2015 entry: positive violations are predicted to produce more favourable outcomes than positive confirmations, and negative violations less favourable outcomes than negative confirmations. Burgoon and Jerold Hale state the same thing in 1988, that positive violations produce more favourable communication outcomes than conformity to expectations. Doing the expected thing is not the best available move.<\/p>\n\n\n\n Their 1988 experiment, in Communication Monographs<\/a>, used friend and stranger pairs, reported with an N of 82, and had one member of each pair significantly increase immediacy, significantly reduce it, or stay at normal levels. Low immediacy produced lower credibility ratings than either high immediacy or conformity, for friends and strangers alike. Low immediacy was read as communicating detachment, nonintimacy, dissimilarity and more dominance than normal; high immediacy expressed the most intimacy, similarity, involvement and dominance.<\/p>\n\n\n\n Two departures of comparable size in opposite directions did not produce mirror image results. That is the point of the word valence. The size of a departure tells you how hard it lands, not which way.<\/p>\n\n\n\n Burgoon’s 1978 paper expanded the 1976 model with explicit definitions and the propositional logic underneath it, then tested five hypotheses. The abstract of that paper<\/a> reports that violations by rewarding communicators produced more positive outcomes than violations by punishing communicators, and that for each type of communicator the relationship between distance and communication outcomes was curvilinear. The same behaviour, from two different people, did not land the same way.<\/p>\n\n\n\n This moderator is the part of the theory most often repeated and most often overstated. The 1989 involvement experiments found that reward mediated only the interpretation of formality, and that high involvement violations produced greater attraction, credibility and persuasiveness than low involvement violations for high reward and low reward communicators alike. In the 1990 study, attractiveness influenced expectancies and evaluations while gender and status had limited effects. Burgoon’s 2015 entry says plainly that some contrary findings have led to revision of the theory.<\/p>\n\n\n\n So the honest version is narrower than the headline version. Who is doing it matters, in some conditions, for some judgements, and the theory’s own author has recorded where it did not hold.<\/p>\n\n\n\n The 2015 entry lists what the theory was expanded to cover: personal space, eye contact, posture, touch, involvement and immediacy violations. Every item on that list is something a body does. The studies match the list. Proxemics in 1976 and 1978. Photographs of people standing at various distances, in various postures, touching in seven different ways, in 1990. Pairs talking for ten minutes and then one person changing how involved they looked and sounded, in 1989. One person in a pair raising or lowering immediacy mid conversation, in 1988.<\/p>\n\n\n\n A chat export has none of that in it. The file is a list of messages. This is a different problem from the one covered in the question of how much a text message leaves out<\/a>, which is about channels going missing. Here the missing thing is a measurement: the theory’s experiments had to observe a body to know a violation had occurred, and the record does not contain a body to observe.<\/p>\n\n\n\n Sort the theory’s own vocabulary against a written record and it splits cleanly. Posture, distance, touch, eye contact and the composite Burgoon calls immediacy are gone. Timing, length, formality and frequency are present, because a thread records when each message was sent, how long it ran, how it was worded and how often someone wrote. Formality is the interesting survivor: it is one of the message interpretation dimensions in the 1989 experiments, and it is the one of those that a written record carries on its face.<\/p>\n\n\n\n Patterns in that surviving column are readable, and other pages here read them: the gap between a message and its reply<\/a>, and the way a polite form can land as distance<\/a>. You can see in a record that someone wrote four words at two in the morning after a week of nothing.<\/p>\n\n\n\n What the record does not hold is the other column. The 1993 design had to install the expectancy before the conversation because the expectancy is not recoverable from the conversation, and the theory’s prediction runs through it. A file shows a departure from a pattern in the file. It does not show what anyone had come to expect, which is the thing a departure is measured against. The nearest related idea, the stock of expectations people build about a stranger as they go, is covered in uncertainty reduction theory<\/a>, and it has the same gap.<\/p>\n\n\n\n A tool reads the column that exists. Portraitor’s own page sets out the two choices it asks for: which one person’s messages are read, with everyone else staying as context, and which time period to cover. That is the behaviour, with the expectations left where they were, which is the shape of reading your own messages back<\/a>.<\/p>\n\n\n\n The claims above come from the University of Arizona’s record of each paper, which carries the published abstract with its authors named. Those are secondary pages. The publishers’ own pages for these articles refused this engine’s requests, so anything in a paper’s body that qualifies its abstract is not reflected here.<\/p>\n\n\n\n The studies are also small and specific. Friend and stranger pairs with an N of 82, fifty one dyads in a ten minute interview, sixty dyads in a problem solving discussion, and a photograph rating task with 622 respondents. None of them was run on a written conversation, and none of the sources read for this page tests the theory on one.<\/p>\n\n\n\nWhere an expectancy comes from before anyone speaks<\/h2>\n\n\n\n
Why a departure gets noticed at all<\/h2>\n\n\n\n
A positive violation is not a small one<\/h2>\n\n\n\n
Who the other person is changes the reading<\/h2>\n\n\n\n
Every experiment behind it was run on bodies in a room<\/h2>\n\n\n\n
What a typed thread still carries of an expectancy<\/h2>\n\n\n\n
The part of this no file can answer<\/h2>\n\n\n\n