Aspect provides information about how events unfold in time. It codes whether events last a relatively short time or a relatively long time, whether events are continuous or repeated, and whether events have finished or not (see Comrie 1976, Frawley 1992). A major distinction is made between perfective and imperfective in linguistic work on aspect. Simply stated, perfective aspect emphasizes the completion or entirety of an event, and imperfective aspect, the ongoing nature of the event (Comrie 1976, Dahl 1985). In describing past events, English speakers typically use the simple past tense form (vERB+ed) in formulating perfective descriptions, as in Roger studied semantics or Maria sold cars, or the past progressive form (was VERB+ing) in formulating imperfective descriptions, as in Roger was studying semantics or Maria was selling cars (see Brinton 1988, Radden and Dirven 2007). In discussing past events, English speakers can also use perfect forms, such as Roger has studied semantics or Maria had sold cars.
Linguists have studied aspect extensively. There is a wealth of information on how aspect develops over time. It is fairly common, for example, for aspectual markers to grammaticalize from lexical items, including motion verbs (see Bybee, Perkins and Pagliuca 1994), and in some cases, from stative verbs (see Carey 1994, for discussion of how English have grammaticalized into a perfect marker). Much is known about how aspectual systems differ across languages. Some languages, such as Russian, for example, make a clear-cut distinction between perfective aspect and imperfective aspect (Bermel 1997) whereas others, such as English, do not (Brinton 1988, Radden and Dirven 2007).
Psycholinguists have paid less attention to aspect, and there are several reasons for this. First, many psycholinguists are interested in pinpointing the mechanisms that underlie language processing, especially sentence comprehension. As such, they focus on the comprehension of sentences as words are being concatenated in real time (see Clark 1997). Because aspect interacts with tense, modality, and other linguistic systems (see Dahl 1985), and because it functions at the level of discourse to some extent (see Hopper 1982), it is challenging to conduct straightforward psycholinguistic investigations. Second, the terminology that is used to characterize aspectual forms is inconsistent. A single aspectual form may be categorized in multiple ways (see Croft 2009, for enlightening discussion). Third, aspect can be marked grammatically and lexically, and this varies cross-linguistically. In English, for instance, a person may say, "I was driving last night," in which the past progressive form temporally extends the event, or "I continued driving last night," in which the word continue temporally extends the event (see Frawley 1992). Fourth, verb semantics partly determine which aspectual form is used and how it is interpreted. For example, imagine that you see a florist accidentally break a vase. In reporting the event later, it would be fine to say, "The florist broke a vase," because break is conceptualized as punctual, but odd to say, "The florist was breaking a vase." Conversely, it would be fine to say, "The florist made a lovely spring bouquet" and "The florist was making a lovely spring bouquet" because make can be construed as ongoing (see Comrie 1976, for comprehensive discussion of aspect and verb semantics).
Of the psycholinguistic work that has investigated aspect, there has been a strong interest in how it constrains the interpretation of situations. Several psychological studies have used narrative understanding tasks to examine how people create situation models. In brief, situation models are imagined "worlds" that can be constructed from processing text or speech (e.g. reading a story), or from memory (e.g. remembering the route you used to take to school as a child, or the layout of a map you studied an hour earlier). These situation models include locations, characters, and objects (see Bower and Morrow 1990). People can imagine different types of motion through space, including motion that is slow or fast, or motion that transpires through a cluttered versus an uncluttered environment. And critically, the way the motion is simulated has consequences for how people recall information about the situation model (Matlock 2004). People can simulate different patterns of movement (e.g. unidirectional path, random), and this alone can influence spatial memory and expectations about future movement (Rapp, Klug and Taylor 2006). People can also update situation models by mentally shifting the locations or positions of objects or characters in a scene (e.g. Morrow, Bower, and Greenspan 1989), and they can readily switch perspective. For example, survey descriptions encourage a bird's eye perspective of a spatial domain, whereas route descriptions encourage a more subjective, ground-level perspective, which is ideal for navigation to a destination (Taylor and Tversky 1996). (For additional information on situation models, see Morrow and Clark 1988, Zwaan, Langston and Graesser 1995, and Zwaan, Maglianoand Graesser 1995.)
In seminal work on aspect and situation models, Magliano and Schleich (2000) used narrative comprehension experiments to investigate how aspect constrains the construction of situation models. Their research focused on how aspect influences the foregrounding and backgrounding of event details. Participants in their study read short passages that contained a critical sentence with a verb phrase marked with imperfective aspect (e.g. was delivering) or perfective aspect (e.g. delivered). Following these critical sentences were three additional statements that reported events that were either concurrent with or subsequent to the situation that was described by the critical sentence. The way people processed the critical situation was probed by measuring the time it took them to verify whether or not a situation (expressed by the critical sentence) appeared in the passage they had read earlier (e.g. deliver baby). These verb phrases were presented immediately after the critical sentence or after the three subsequent sentences. The results showed that after reading the critical sentence and three subsequent sentences, people were quicker to identify the verb phrases that had been in the prior text when those phrases had been marked with imperfective aspect (versus perfective). Their findings suggest that imperfective aspect can increase the prominence of an action (more foregrounding) more than perfective aspect even though the event was objectively the same. (For related work, see Carreiras, Carrido, Alonso and Fernández 1997.)
In other groundbreaking research, Madden and Zwaan (2003) investigated the way aspect constrains the understanding of events in situation models. Participants in their study viewed pictures of events that appeared to be in progress or that appeared to have just completed. Participants then had to indicate whether the pictures matched verbal descriptions that included imperfective or perfective aspect. For instance, participants viewed a picture of a person kneeling next to a fireplace, in which the person is still building the fire or has just ignited the fire. Then they had to decide whether accompanying descriptions such as "made fire" (perfective) or "was making a fire" (imperfective) matched. On average, participants were quicker to match pictures of completed actions (versus incomplete actions) with perfective descriptions, but no slower or quicker to match pictures of completed actions (versus incomplete actions) with imperfective descriptions. In brief, these results suggest that imperfective aspect constrains the understanding of a situation by encouraging the reader to take an internal perspective, and as such, it enables greater attention to details of actions. In contrast, perfective aspect gives an external viewpoint of a situation, and encourages focus on the end state of the situation. (Related work is reported in Madden and Therriault 2009.)
In other pioneering behavioral work on aspect and situation models, Morrow (1985) explored how imperfective and perfective descriptions of motion events affect how people conceptualize movement through imagined scenes. Participants in the experiment had to study the layout of the rooms in a house, and then read a short passage about a person moving from a Source location to a Goal location in the house. The sentences in the passage included a translational motion verb (e.g. walk) marked with either imperfective or perfective aspect as well as a Source location (e.g. kitchen) and a Goal location (e.g. bedroom), as in John was walking from the kitchen to the bedroom or John walked from the kitchen to the bedroom. Participants often located the character described as moving on the path somewhere between the Source room and Goal locations after reading imperfective motion statements, but in the Goal room after reading perfective motion statements. The results suggest that imperfective aspect draws attention to the unfolding details of a situation, whereas the perfective aspect draws attention to the terminus or resulting phases of a situation.
Anderson, Matlock, Fausey and Spivey (2008) further investigated the role of aspect in conceptualizing motion events but they introduced a method that allowed them to pinpoint where and how motion transpires. They used a (computer) mouse-movement study to examine movement along a path in response to either imperfective or perfective verb phrases. In the study, participants were shown a large picture of a path on the computer screen. The path started at the lower part of the screen, and ended at a destination (e.g. a school, hospital, park) on the same screen, and next to the picture was a small static silhouette character, for instance, a man who appeared to be jogging (e.g, slightly bent leg and arm in front, slightly bent leg and arm in back). Participants heard a sentence that described the character moving and arriving at the destination (e.g. Tom was jogging to the woods and then stretched when he got there [imperfective] and Tom jogged to the woods and then stretched when he got there [perfective]). All imperfective and perfective versions of this sentence included translational motion verbs, such as jog, ride, and hike and a to + location phrase about the destination. As soon as participants heard the description, they clicked on the character and placed it in the scene to match the description they heard. On average, participants moved the character along the path toward the destination more slowly with imperfective motion descriptions. Similar results were obtained in a follow-up study by Anderson, Matlock and Spivey (2010) with improved stimuli and a broader range of sentences and situations. The results of these studies suggest that imperfective aspect reflects greater attention to the ongoing process of motion toward a destination.
The behavioral studies summarized above provide good insights into how aspect constrains the way people conceptualize events in the situation models they construct and update. In particular, imperfective descriptions encourage an internal viewpoint by drawing attention to the ongoing state of events, at least more than perfective descriptions do (see Madden and Ferretti 2009 for additional discussion). Because these results are consistent with the aspectual patterns that linguists have observed in many languages, they may initially seem unremarkable. Semanticists know, for example, that imperfective aspect expands the temporal window of a situation because it is associated with unbounded, ongoing events in its basic construal (see Frawley 1992, Radden and Dirven 2007, Talmy 2000). From this, it follows that people might infer more time permits more action. Still, what happens when the time periods in imperfective and perfective descriptions are identical, as in John was reading for an hour versus John read for an hour? Is more action still inferred with the imperfective? Such questions are worth investigating because they may lead to even deeper insights into how people produce and understand aspect in everyday language.
Recent work on aspect investigated this very issue. In a study by Matlock (2011), participants did a sentence completion task. They completed a sentence that began with one of two adverbial clauses, either "When John walked to school" (perfective) or "When John was walking to school" (imperfective). On average, participants mentioned more actions in their main clauses when framed with imperfective information (e.g. When John was walking to school, he felt sick and went home) versus perfective (e.g. When John walked to school, he got a hamburger on the way). In a second study, on aspect and telic verbs, participants read the statement, John was painting houses last summer or John painted houses last summer, and answered the question, "How many houses?" On average, they estimated more houses were painted with the imperfective statement. In a third experiment, on aspect and atelic verbs, participants read the statement, John was driving last weekend or John drove last weekend, and answered the question, "How many hours?" Overall, they provided longer driving time estimates in response to the imperfective statement. The results of these experiments suggest that more action is conceptualized in a given time period with imperfective aspect.
This current work further explores the role of aspect in the interpretation of event descriptions. It is known that one of the main jobs of aspect is to establish a temporal window in which a set of actions or states occurs (Li and Shirai 2000). Clearly, this is important. However, it is also useful to consider other ways that aspect contributes to everyday language processing, including how it shapes inferences about type and amount of action in a given time period. It is also important to explore how aspect can shape thought and communication in natural discourse. Toward this end, we constructed an experiment that resembles a police interview after witnessing a car crash. Participants in our study were shown video clips of vehicular accidents and asked to report what was happening (imperfective framing) or what happened? (perfective framing). Their responses were analyzed for speech content, including number of motion verbs and reckless driving phrases, and gesture content, including number of iconic gestures which are depictive of actions and other key elements in descriptions.
-  Telic verbs imply a goal and end state. Atelic verbs do not.