User Actions as Click Acts

In contrast to the logical positivists of the Vienna Cycle, Austin argued that the meaning of an utterance cannot be reduced to its verifiability (Saeed 207). In Austin's opinion, utterances do not only have the purpose of making statements but of performing actions (1-5). Furthermore, even statements can be regarded as actions (Saeed 210).

 

Figure 8: A MacOS Contextual Menu

It is not surprising that all user utterances examined in the previous section have to be regarded as actions. It is doubtable if there are any user utterances which could be classified as statements at all.⁵¹
Austin distinguishes between explicit and implicit performatives. An explicit performative in human-human communication such as

I invite you to come up and see me sometime (Example taken from Saeed 210),

includes what Austin calls a performative ("I invite"), whereas the utterance

Come up and see me sometime (Example taken from Saeed 210)

does not include such a performative. It could be argued that there are only implicit performatives in human-computer interaction, since the user does not have to say "I (hereby) order you to cut the selected object" (although the user is free to have this sentence in mind). As pointed out by Saeed, most performatives in human-human communication are implicit, so that it could be assumed that the permanent use of explicit performatives would be perceived as unnatural by the user (210; see also Gramley and Pätzold 209). More importantly, in regard to what has been said in section 4.1.3, it would be difficult to find the appropriate performative verb. In the example above, the performative "order" was used, but it could also have been a request.
Austin does not only emphasize that utterances have to be considered as actions; he also distinguishes between three elements of a speech act,

• a locutionary act, which is the actual utterance,
• a illocutionary act, which is the action intended by the speaker,
• and a perlocutionary act, which is the effect of the illocution on the hearer (Saeed 212).

By uttering "It's cold in here" (locutionary act), a speaker might want a hearer to close the window (illocutionary act), and might receive the answer "Yes, it is really cold in here". The hearer might have understood the illocutionary force but is reluctant to get up since the speaker is nearer to the window, which could be regarded as a perlocutionary act. Or, the hearer has not understood the illocutionary force, but is polite enough to answer the speaker's utterance. In the latter case, the illocutionary force has not reached the hearer. Since the perlocutionary effect is of less interest to linguists, the illocutionary act is often meant when referring to speech acts (Saeed 212; Yule 49).
While the hearer in the example above has to decode the illocutionary force, this is not the case when an utterance is made to a computer system, since, as pointed out in the previous section, most of the utterances and the functions behind it are predefined. Nevertheless, it is possible to apply Austin's speech act theory to human-computer interaction; a "click act", as it will be called in the following, consists of three elements:

• the locutionary act: user input as described in the previous section
• the illocutionary act: the action intended by the user
• the perlocutionary act: the action which takes place after the mouse click, in most cases a function provided by the programmer of the object

It can be concluded that the locutionary act is the same as the perlocutionary act in most cases; if a user selects the operation "Cut" from the menu, only a system error could prevent the system from cutting the desired object. In contrast to this, the illocutionary act may be totally different from the perlocutionary act. On the Macintosh, when dragging a file from one disk icon to another, it is copied but not moved as one might expect, for example.
Similar to Austin's explicit and implicit performatives, Searle distinguished between direct and indirect speech acts. The difference between these approaches is that Searle defines a direct speech act as the conventionally expected function and the indirect speech act as an extra actual function (Saeed 214). Thus, the sentence

Why don't you finish your drink and leave?

includes a direct act, a question, and an indirect act, a request (example taken from Saeed).
All click acts seem to be direct acts since there is no possibility for users to utter indirect acts. Also, it is questionable that users have an indirect act such as Would you mind cutting the selected object? in their heads when selecting "Cut" from the menu bar.