As Cohen , p. Nondiegetic background music has many functions in movies, but at least two are relevant here.
First, even more than lighting, music sets a mood. It likely does this through associative processes that viewers share about the particular music: its key major or minor , its genre classical, popular, ethnic , its relative tempo, and its relative familiarity.
Second, music often provides advance information about an upcoming event or change in the plot, and indeed Boltz, Schulkind, and Kantra ; Magliano, Dijkstra, and Zwaan ; and S.
Tan, Spackman, and Bezdek have shown that movie viewers are sensitive to this fact. Filmmakers often use this foreshadowing so that viewers can be ready for what happens next, and this can greatly aid processing and memory. For these reasons, one can expect that filmmakers use nondiegetic music at points in the narration when important events and turns are about to happen. But does this too have a pattern across movies? Most of the music in the smaller sample was nondiegetic.
Moreover, they found no apparent trends over the period of movies investigated here, from to Thus, one should expect that the median use of music in the smaller sample of movies would be in a bit less than half the shots.
As before, each movie was normalized into equal duration bins. No other normalization took place. Results are shown in the left panel of Fig. The distributions of music Study 5 and of shot scale Study 6 in 23 movies across the duration-normalized time bins. The pattern for music as it is distributed across movies is different than for any of the previous variables.
One might think that there would be a difference here between movies with trimmed versus overlaid credits, and there is but in a surprising direction. Because the seven movies in the former group are all older movies, this may reflect stylistic differences that have changed with time. Over the remaining bulk of the movie, the music gradually and linearly increases to accompany over half of all shots.
I fit these later data with a linear regression so show its sharp contrast with the prolog. In other words, after an initial flurry of music at the beginning of the setup—again, more strong evidence for the existence of a prolog—filmmakers drop back and gradually build up the nondiegetic music as the narration progresses.
Likely they do this in part to increase tension and further induce viewers to remain transported by the narrative. This more-or-less uniform trend would also suggest that music is bound to the fabula of the whole movie, not to that of the act structure of narrative.
Nonetheless, it also likely reflects the emotional content in separate scenes scattered throughout the different narratives, and that the scenes get linearly more intense across the narration. I will suggest later in the discussion of Study 6 that the organization of individual scenes is also quite independent of the four acts, and that averaging across scenes creates no special pattern in the act structure of movies.
The close-up has inspired fascination, love, horror, empathy, pain, unease. It has been seen as the vehicle of the star, the privileged receptacle of affect, of passion, the guarantee of the cinema's status as a universal language. The use of different shot scales—a close-up, a long shot, or something in between—is a staging decision within film style. The motivating question then is: Is there variation in shot scales across the larger units of the narrative? Using the shot scale data of Cutting, Brunick, and Candan , I divided the movies in the smaller sample into bins, determined the median shot scale within each bin for each movie, and then averaged these data across movies.
Unlike in other studies, I did not normalize the data, in part to compare them with previous work. Results are shown in the right panel of Fig. Notice that except for the first three bins there is no patterned change in the median shot scale across the rest of the movies, and even this inflection is surely because every movie must have a beginning of a first scene, which starts with Bin 1.
Thus, whatever information about scene change is carried by shot scale, filmmakers use none of it to mark the changes across acts. Furthermore, the long flat section of the scale values again implies an independence of film form for acts and scenes.
With respect to any theory of the larger narrative, the focus of this article, shot scale is essentially unvarying. However, at a more local level—the level of the scenes within the narrative—the pattern is much different. On average the movies in this sample have a new scene or subscene every 55 seconds. The insert in the right panel of Fig. Scenes tend to begin with a longer-scale shot and become shorter as the camera moves in to show more of the faces of the characters, and with a lengthening at the end of the scene when the cinematography often backs off before a cut to a new scene.
Cutting, Brunick, and Candan found that shot scale is the most important cue for scene change, followed by transition type cuts vs. That the latter two provide patterned information about act structure suggests that different physical parameters of movies are associated with different-sized units of narrative structure. Music and shot scale offer no additional information about the act structure of movies.
The use of music appears to be conditioned on the stylistic decisions of the filmmakers at the level of whole-movie narration, building incrementally throughout, whereas the use of shot scale is a function of stylistic decisions and the level of the scene, which then average out as the succession of scenes proceeds at different rates in different movies.
The general independence of these levels, I would contend, adds flexibility to narration. Among the three proposed here, one source of information tugs at the attention of the viewer in the domain of 2 hours or so music , others in the domains of about a half-hour or less shot duration, shot transitions, motion, and luminance , and still another in the domain of a minute or so shot scale. Moreover, both music and shot scale provide continuing support for the prolog—the amount of nondiegetic music is greater, the shot scale is longer, and, looking back to Fig.
The evidence is sufficiently strong for a prolog within the setup that I have added it to Fig. The previous six studies have dealt with narrational information from the shots of movies—durations, transitions, motion, luminance, music, and shot scale.
The next four deal with aspects closer to the narrative proper—determining when characters are introduced, when characters talk to one another, when genres might diverge, and when scenes change. In all the manga that I have read and anime I have watched, the protagonist always appears in the first chapter of manga or first episode of an anime. The defining characteristic of the setup is the introduction of major characters, but when exactly do characters first appear?
I will call most of these the other characters , plus add a few that are important to the plot of a few movies. From these I single out the protagonist or protagonists one to three. In the same way I have divided these movies before, I noted when each character first appeared and assigned that appearance to the appropriate bin. I normalized the bin entries so that together they summed to 1. Thus, in a movie with 10 characters, the appearance of each counted 0.
Lead character were done in the same way. For a movie with three protagonists, each would count. Again, each movie contributed equally to the results. For comparison, I plotted the appearance of new locations in each film as well, and assigned them to the appropriate bins. The average movie in this sample had 31 locations, so weights per bin were adjusted accordingly. Unlike other analyses, however, I accumulate the results until they added to 1.
Moreover, those few who had not been introduced were among the movies with more than one protagonist. Such early appearance is not true for locations. Just under half of the locations have been seen during the setup, and these are then introduced fairly uniformly throughout the rest of the movie. Thus, and again, there is a heavy emphasis on a difference in the structure of the movie within the first 5 to 7 minutes; the prolog typically introduces the protagonist s.
Notice that there are often 10 to 20 minutes remaining in the setup after the introduction of a protagonist. Clearly, this is when backstory and exposition occur as well as the introduction of subsidiary characters. Following Gernsbacher , this is when the foundation of the narrative is laid. Finally, notice that the setup is not strongly identified with the introduction of locations. Music is not the only important content in the audio track; indeed, speech is even more dominant.
Cutting and Candan categorized every shot in the smaller sample of movies into 15 different types of shots. A variant of this is the over-the-shoulder shot, where the camera is behind one character whose back is turned and is focused on the other character, who is talking. A third can be either over-the-shoulder or without the turned-away character but with the character facing the camera not talking, but listening to the other.
This is called a reaction shot. Finally, there are occasional conversational shots over telephones, intercoms, or holographic devices in science fiction movies that are staged in a similar way to regular conversations. All conversational shots were apportioned to the bins according to the lenght of the movie, and coded for the presence 1 or absence 0. These were then averaged within bins and the mean proportion per bin is shown in the left panel of Fig.
Vertical lines separate the setup, complication, development, and climax as in Figs. Notice three trends in the data. Moreover and again, there is no difference between movies with and without covering credit sequences.
Second, the conversational density is quite constant but noisily varied through the complication and development.
Third, conversations decline during the initial portion of the climax, but rise again in the epilog. The decline in conversations at the beginning of the climax suggests that it is finally time for the protagonist s to act, and not to talk. But again, the most striking trend is in the prolog, where few conversations occur and gradually increase over the first two-thirds of the setup.
But other shots may be more distinctive. As suggested by Bordwell in the epigram, perhaps the most salient—and the type that most clearly separates genres—is the action shot. If one is looking for a general formula for movies, surely one of the best ways to test this idea is to look for a shot type that distinguishes genres and may rupture the idea of a uniform movie format.
It is no ordinary combat. The fighters leap twenty feet in the air, pivoting and somersaulting, sometimes clashing with one another. Movies, like life, are basically about talking and doing. In their taxonomy of shot types Cutting and Candan defined an action shot as one with beyond-normal physical activity—not only those in fights of all kinds but also those in sports, accidents, explosions, chases, building collapses, and other more or less extreme events. Helpfully for classification, in sequences these shots are almost always covered by nondiegetic music.
Not all shots in action films are action shots, but many are. Moreover, and equally important, a number of action shots are found in films that are not action movies.
Unsurprisingly, action shots in dramas are rare 1. Indeed, there were none in three of the eight dramas in the smaller sample. Nonetheless, action shots in comedies are more prevalent 6. Thus, it is worth comparing action shots in action movies versus those in the other films. Because action shots are relatively sparse in most of these movies, I divided the movies into 20 time bins, found the proportion of action shots within each bin, and then averaged those values across movies.
The data for the seven action films and the 10 remaining comedies and dramas separately are shown in the right panel of Fig. Notice the two humps, one at the boundary between the setup and the complication and the other just into the climax. The latter makes considerable sense—there should be a rise to higher activity generating more action shots, in the climax, and these should fall off during the epilog.
The first hump, on the other hand, was a surprise. Action shots are just as prevalent here as in the later cluster and suggest that part of the inciting incident or the lock-in at the boundary between the setup and complication in may often be preceded and followed by strenuous activity.
Although the data are noisy, again the action films show a similar pattern but with greater frequency and an interaction showing increasing frequency across the runtime of the movie. One possible reason for the relative weakness of the higher order polynomial fit was mentioned earlier in the context of Study 3— action sequences can occur at almost any point in the runtime of an action movie, and the relative noise in these data bear this point out.
Another is that there are only seven action movies sampled. Nonetheless, I find the general parallel between the fourth-order fits to the action and nonaction films worth preserving, suggesting that all genres have more or less the same structure with the pattern in action films considerably ramped up.
The uptick in action shots at the boundary between the setup and the complication also serves to further distinguish those two acts, and given some substance to the notion of a lock-in, the turning point between the two acts. If character introduction and laying the foundation of the narrative is the function of the setup, and if conversational exchange is the function of the complication and the development, action is clearly the function of the first part of the climax.
How then do these facts affect the scene structure of the narrative? In drama, scene refers to a division within an act of a play, indicated by a change of locale, abrupt shift in time, or the entrance or exit of a major character. Let me now shift to discussion of the scene. Continuity and discontinuity, the psychological impression in movies of ongoingness versus change, are not what they might first appear. Strikingly, despite the glaring and abrupt physical differences, cuts do not always disrupt the continuity of the unfolding story see, e.
By this formulation there are seven kinds of narrative shifts generated by the presence or absence of a shift along the three dimensions. Shifts and nonshifts at shot transitions yield 2 3 or eight possibilities, but when all three dimensions do not shift the result is continuity. This fact subtracts one out, yielding seven. Shifts in location, regardless of whether or not there is a shift in one of the other dimensions, are quite common, occurring in the smaller sample an average of 98 times per movie.
Shifts in characters are a bit more prevalent, averaging per movie, but shifts in time are less so, averaging only 49 per movie. Again, typically more than one dimension shifts at a time so the average number of narrative shifts of one kind or another in this sample of movies is Cutting, How are narrative shifts distributed across the length of movies?
Cutting, Brunick, and Candan had viewers segment 24 movies into scenes, three viewers per film. Each viewer marked the frame number where a new scene began. Analyses below are carried out on the three types of narrative shifts—location shifts whether or not characters or time changed, character shifts whether or not location and time changed, and time shifts whether or not locations and characters changed. As I did above for noncuts and because these data are sparser than those of other studies, I divided the 23 movies in the smaller sample into 20 equal length time bins, recorded the number of narrative shifts in each, and normalized them to sum to 1.
Narrative shifts of all three types are considered together in the upper left panel of Fig. Given that shorter shot durations and more motion occur during the climax this should not be a surprise.
Note that again the first bin the first 5 to 7 min of a movie is an outlier, with many more scene or subscene changes than at any other point in the movie. And again, this is evidence for a separate prolog within the setup.
This is likely because the typical movie often dodges around to different locations with different characters in an exposition phase of the narrative.
Normalized density distributions across 20 consecutive, duration-normalized bins in 23 movies, for all narrative shifts, and all changes in location, in characters, and in time Study The general shape of the function promotes an implication.
Given that there are fewer narrative shifts in the middle of a movie, those scenes must be longer than elsewhere. Indeed, the length of scenes and subscenes in the setup and climax averages This trend seems allied with the results of Study 8, which showed that the middle of a movie has many more conversations than at either end. Conversational scenes are longer. As shown in the lower panels of Fig.
Since these two types of shifts are the most common and most correlated it is not a surprise that the summary shift pattern in the upper left panel is essentially the same. The data and quadratic fit for locations are a bit more lopsided than that for characters, with location changes more frontloaded in the narrative. As shown in the upper right panel of Fig.
In other words, time shifts become fewer across the narration. This result makes sense as the protagonist moves toward achieving her goal.
Diegetic time becomes more like real time. It is interesting to consider two types of time as discussed in the narratology of literature, and how these become particularly acute in movies. This comparison concerns narrative speed see, e. Paxson , p. As Paxson and Genette define them summaries in movies are rare.
They typically occur in voiceovers often in the prolog as in All About Eve, or in text-overs often in the epilog as in Erin Brockovich, , and The Social Network, Pauses are also rare but can occur when the cinematography freezes frame while a narrator continues, as happens in All About Eve and Goodfellas The data of Fig.
This undoubtedly gives the viewer a sense of urgency as the narrative increasingly proceeds in real time as it moves toward and into the climax. That is, a goal is set and must be attained by the protagonist at a particular time. Movie narratives, as they approach that deadline, appear to have a strong tendency to proceed with the coupling of diegetic narrative time and real time narration time. Action films particularly have this feature e. Study 7 affirmed the functionality of the setup.
Most characters, and particularly the protagonist s , are introduced and their goals become known. As Gernsbacher would note, this establishes a foundation on which the relations among subsequent events can be laid. Study 8 continues the thread of functionality, showing that the setup is relatively devoid of conversations but that the complication and development are rife with them.
Clearly, these conversations carry the bulk of the narrative progression and conflict within the plot. Study 8 also showed that conversations become less prevalent at the beginning of the climax, when the protagonist is driven to action, but their frequency returns for the epilog. Study 9 follows the most physical activity in the narration, finding in dramas and comedies that it peaks at the end of the setup and again at the beginning of the climax—both times when conversations are less.
It also showed this pattern for action films but with an additional underlying increase throughout. And Study 10 plots the frequency of scenes and the changes between them. Scenes are longer in the complication and development, likely because of the conversations, and shorter in the setup and climax.
Location and character changes across scenes follow this pattern, but time changes do not. Instead, time changes become incrementally less frequent. That is, narration time and narrative time converge, likely because of the complications of story structure around a deadline that must be met. The previous 10 studies have each taken single attributes movie narration and explored how those unfold over the course of the average film. Their results provided useful information, much of which validated a four-act narrative theory of popular film.
However, it seems logically implausible that these results would reflect ten independent dimensions of movies. Instead, since the structure of the world around us is a high-dimensional but vastly intercorrelated space Edelman, , movie structure should be as well.
Thus, the last two studies seek to place these attributes in a lower dimensional framework. Simply put, cognitivists like film. We are interested in understanding the basic processes of how movies work: how they are structured, how they convey meaning and evoke emotion.
In the previous 10 studies I analyzed movies in 10 different ways. Some connections became obvious, as in Studies 1 and 2, with the strong negative correlation between transitions of all kinds and noncut transitions. Other oppositions, however, are less obvious.
One way to explore these is through principal component analysis. This statistical technique converts different sets of matched data into a set of linearly uncorrelated variables called principal components.
Given 10 inputs, one could search for as many as 10 components, but this would be unhelpful. Instead, I will search for a reasonable solution with as few components as seems warranted, and then search for dimensions of film style within them. The median bin values from each study were entered as columns in a larger data set. I used the raw medians for the bins from most studies 1, 3, 4, 5, 6, and 7 , and I interpolated values to achieve estimates from those studies with only 20 bin medians 2, 9, and Footnote 11 That is, for example, what were Bins 1 and 2 became Bins 1 and 6, with intermediate bins having values linearly interpolated in between.
This expansion procedure was continued Bins 2 and 3 became Bins 6 and 11; 4 and 5 became 16 and 21, etc. With the 10 normalized inputs I then ran the analysis. These results can be most easily visualized in a loading plot of the two components. Following Pythagoras, the maximum variance accounted for across two components is the square root of the sum of the squares of component correlations for each variable.
Thus, the results can handily be plotted in a circle. Such a loading plot of the 10 variables is shown in the left panel of Fig. Notice that all but one of the data vectors, brightness from Study 4, approaches the circular limit.
The average length of the nine longest vectors is. This means that the data of nine of the 10 studies is well captured by these two components. The left panel shows a loading plot of the first two principal components and the vector strengths of the data in those components for the variables from each of the 10 preceding studies.
The right panel leaving out brightness provides a dimensional interpretation of these data as four correlated, opponent continua. As expected the vectors for Studies 1 and 2, transition densities and noncut densities, are opposed and can be combined, as shown in red in the right panel of Fig.
But notice that other oppositions are nearly as prominent. This makes sense given that diegetic music is rarely heard in the background of conversations and that nondiegetic music often occurs when no character is talking. This too makes sense. That is, shorter scales—close-ups and medium shots—are used in the middle of scenes, whereas longer scale shots tend to be used at the beginnings of scenes, the time when new characters are typically introduced.
Luminance, shown in light blue in the left panel, is ill fitting in this plot. These patterns allow for a set of interpretations, shown in the right panel, of four correlated bidirectional dimensions in the data.
This type of representation generated from a loading plot is nonstandard, so let me work through it. The dimension whose component vectors fit best within the loading plot is shown in red, and it is strongly related to the first component. I will call this the editing dimension. The second best fitting dimension, is in black, and is most closely aligned with the second component. It combines motion and action shots, and I will call it the motion dimension. The third dimension, in green, pools music and conversations.
I will call it the sound dimension, and it cuts diagonally across the two components. The fourth dimension concerns framing and combines three vectors.
It pits the values of shot scales with character introduction and scene changes. This dimension is also closely aligned with the first component. And the one dimension not shown and strongly represented in a third component is lighting. What is particularly pleasing about this arrangement are the cross-dimensional correlations that they suggest, replicating many findings elsewhere.
The association between shot scaling and duration was broached by Bordwell and explored and supported in detail by Cutting and Cutting and Armstrong This is valid because conversations are typically filmed with medium close-ups and without music. In this manner, a principal component analysis allowed the derivation of four correlated conceptual dimensions of movie narration.
These are laid out across two components and shown in the left panel of Fig. Moreover, these dimensions seem to capture well the relations among nine of the physical dimensions explored in the studies presented here. So far, of course, this analysis ignores the time course of the narration, but more can be done within this componential framework. In particular, the dimensional analysis of Fig.
Narration is more than an armory of devices; it becomes our access, moment by moment, to the unfolding story. Narration in any medium can usefully be thought of as governing our trajectory through the narrative. Within the bin framework, I consider Bins 1—3 to correspond to the prolog, Bins 4—25 to the setup minus the prolog, Bins 26—50 to the complication, Bins 51—75 to the development, Bins 76—97 to the climax minus the epilog, and Bins 98— to the epilog. Given that several of the five normalized dimensions of Fig.
I then calculated the median value of each dimension black dots and the interquartile range red bars across each of the six parts of the narrative and fit these back onto the configuration in Fig. The results are shown in the six panels of Fig. In addition, a filled green circle at the intersection of vectors in each panel represents the relative median amount of luminance in each section.
Distributions of values on five movie dimensions as in Fig. The extent of each bidirectional arrow corresponds to the 90th and 10th percentile of values in those domains, the black dot corresponds to the median value of each narrative part on that dimension, and the red bars correspond to the interquartile range.
When black dots and red bars cover the gray arrow at the ends of the dimension, this means that those values extend beyond the 90th or 10th percentile. The size area of the green dots in the middle of each display corresponds to the luminance in the third principal component. The results show in a graphic manner the differences across the sequential parts of movie narration. The prolog has extreme values on three dimensions—a lot of nondiegetic music, long duration and long scaled shots—with varying motion and luminance.
The values for the setup without the prolog generally converge to the center of the configuration except that the images darken only a bit and the nondiegetic music nearly drops out. The complication has a bit more music, and with shorter scaled and shorter duration shots and average brightness. Only motion remains roughly as it was in the setup, but even here it extends into the higher range.
The development swings back on two of these dimensions—a bit less conversation, a bit longer shot durations, but also becomes darker, and keeping shot scales short while maintaining roughly the same amount of motion. The climax without the epilog adopts extreme values of elevated motion and short shot durations, maintains a short shot scale, has much more music and fewer conversations, and becomes markedly brighter.
Finally, the epilog is bright, swings back markedly to very little motion and to long duration shots, but to a midrange of shot scales and a mix of conversation and music. Across the first 10 studies, I measured samples of films in 10 different ways accruing evidence from the syuzhets for a large-scale formula in the fabula of popular movies that has persisted for at least 70 years. In addition, a larger scale whole-film set of patterns appears for nondiegetic music, and for shot scale in the smear of different scene onsets across many movies, as suggested in the results of Studies 5 and This medium-scale formula—actually a theory of popular movie narratives—is one that states that the fabulas of movies generally have four acts, as suggested by Thompson and Bordwell , some with optional subdivisions within them.
Basically, the formula is halfway between Aristotle and Horace—four acts of roughly equal duration, not three or five. These acts are the setup, the complication, the development, and the climax, with a likely prolog and epilog within the first and last acts, respectively. The fact that Studies 1, 2, and 4 find results that distinguish between bins in the complication and the development suggest that a three-act theory Field, is insufficiently fine grained.
The fact that no theory with more than four acts has specified any time frame on those acts makes them untestable in the manner that I have approached them.
And the fact that the combined results in these studies are consistent with a theory proposing four acts with roughly equal durations suggests that it was amply tested. Let me review the evidence for this four-act structure and the likely psychological impact of those narrational measures on the viewer.
The existence of a prolog within the setup is not part of any narrative theory of film that I know. It has been hiding, if not in plain sight, surely behind the ubiquitous opening credit sequence that has been used for more than 60 years. It is strikingly salient in almost every domain of data that I have investigated. That is, the first 5 to 7 min of narration one, two, or three bins of the hundredfold divisions, or the first bin of the twentyfold divisions are often markedly different than those that immediately follow.
On average, shot durations are much longer Study 1 , noncuts more frequent Study 2 , nondiegetic music more prevalent Study 5 , shot scales longer Study 6 , more characters introduced Study 7 , fewer conversations heard Study 8 , and with more narrative shifts Study 10 than anywhere else in the movie. Nonetheless despite its unique characteristics, like an epilog, not every movie has a prolog. The existence of a prolog seems not to be an artifact of having opening credits superimposed on early shots.
The pattern of shot durations in Study 1 is the same for those movies with and without overlaid credits; remember, the older movies with text-only title-card credits had these removed before the binning operation. Motion patterns are the same in both groups of films. The pattern of nondiegetic music also seems not related to the presence or absence of credits. My guess is that filmmakers introduced overlaid credits knowing that early shots of a films were already necessarily slow, and that placing credits over them solved the dual problem of familiarizing viewers with the locale and tenor of the story while also satisfying the need of putting the names of the major stakeholders up front.
Many aspects of popular filmmaking are about optimizing resources, and overlaid credits with the normal content of the prolog allows a two-for-one fulfillment of needs. It is the portion of the movie in which the filmmakers write their contract with the viewer. That contract states that what follows will be an absorbing and interesting story, and the viewers effectively sign on to be involved Cutting, Thompson , Field , and others agree: the setup including the prolog introduces the characters and their issues.
Study 7 provided strong evidence that this is the case. This rapid character introduction differs sharply from the first appearance of the various locations in movies, which are distributed more evenly through the narrative. Beyond character introduction, the setup is separated from the complication by average shot duration Study 1. By the end of the setup action shots, regardless of genre, may lift the physiological involvement of the viewer Study 9.
But basically, the viewer has dug into the narrative—with cognitive and perhaps a bit of physiological anticipation.
She may have experienced the first plot point the inciting action and awaits the first turning point the lock-in , trying to read the mind of the protagonist as to what she will do next.
These acts are the domain of conversation Study 8. The complication is markedly separate from the setup in character introduction Study 7 , scenes and subscenes begin to lengthen Study 10 , and the narrative begins to take off in a new direction after the protagonist has realized that her initial goals are thwarted.
Basically, it appears that these measures are indications that the story is getting more complex as it passes the midpoint. Evidence is strong for the climax as different from the development, and for the epilog as different from the rest of the climax. Conversation diminishes Study 8 , scenes and subscenes begin to shorten and narration time and narrative time converge Study 10 , shot durations plummet Study 1 , noncut transitions become fewer Study 2 , motion surges its highest levels Study 3 , brightness returns Study 4 , and nondiegetic music is at its most frequent since the prolog Study 5.
In viewers, heart rate and skin conductance should rise, and viewer absorption in the movie is likely to be at its peak. In the epilog shot durations then become longer than at any time since the prolog; dissolves and conversations may return; brightening continues. All is well, and cardiographic and electrodermal readings should begin to approach normal levels. The viewer can leave the theater much as she entered it, but having had an emotional ride and some intellectual exercise in theory of mind.
In the final two studies, through componential analysis, I condensed the variables of the 10 earlier studies to five dimensions based on film style. Filmmakers craft the syuzhet through decisions about staging, framing, editing, sound, focus, and color.
However, I did investigate here the other aspects of film style: staging Studies 3, 4, 7, and 8 , framing Study 6 , editing Studies 1, 2, 9, and 10 , and sound Studies 5 and 8. As it turned out, leaving luminance aside, rather than nine dimensions for these measurements of movies there are really only four, and even these are often strongly correlated. One dimension concerns editing shot durations, short vs.
Another dimension lighting is not strongly correlated with these four. Together, these five dimensions encompass central aspects of the basic toolkit available to filmmakers, who clearly make good and thorough use of it in crafting the structure of the narration to reflect larger narrative units. Values on these dimensions dance around considerably as the narration progresses, as shown in Fig.
One might accept these results but wonder what good they might be in psychological terms. First and most broadly, these results address an aspect of a decades-old issue in cognitive psychology; we talk about information processing but we often have no real idea how rich the information in the stimulus actually is; that is, what information is available to be processed. Although these studies measure no viewer responses, they do reveal variations in sources of information that have been shown to affect viewers.
This is important because movies and movie clips have been, and increasingly are, used as stimuli in psychological studies—of emotion Hutcherson et al. Moreover, psychologists are becoming increasingly interested in whole movies as a test bed for understanding perception, cognition, emotion, and other psychological responses more broadly Kaufman et al. Second, the normative data here provide strong evidence that the narration the surface form of the narrative, or the syuzhet typically contains many physical cues, continuously distributed, that likely guide attention and arousal and that could guide movie comprehension.
In the terms of McNamara and Magliano , p. This is important because, against some views of language structure, I would claim that aspects of form and meaning in popular movies are not independent. That is, there are strong correlations between the progression of the narration and the narrative states of movies. Third, because psychological studies of text have focused on processing and comprehension, they have been confined to relatively short stories. Subjects can sit through only so much.
To be sure, a great deal has been learned in this domain, but the structures of short stories are likely more impoverished than those of longer ones.
Thus, the opportunity for understanding larger scale structures in textual narratives and how they might be processed may have been missed. Neither processing nor aspects of comprehension were studied here, but this analysis of movies allowed for the empirical explication of larger scale narrative structures in stories that may prove fruitful in work on discourse processes.
Finally, I would hope to reinvigorate the psychological study of multiple sources of information and preferably across dynamic situations, as suggested in Fig.
Most of our interactions in the world are surely guided by multiple information sources from which we select and combine cues. In the introduction I reviewed the proposed theoretical structure of plays and films. These were in terms of acts and turning points, and my analysis throughout this article has been fairly tightfisted in distinguishing among theories. Darwin, , p. Nonetheless, let me now swing the other way, become a lumper, and explore the similarities in narrative structure across the stories of different media.
To be sure, the exact ordering and segmentation of some of the parts shifted across different theoretical views. But more broadly there was great similarity, and that by itself is evidence for an art-form general and formulaic narrative structure. Of course plays and movies are not the only narrative forms following this kind of scheme.
Labov and Waletzky analyzed short oral histories and life narratives given by inner-city adults and gleaned a structure that included six sections. The first, the abstract , is perhaps like a prolog—optional, short, and declaring what the story is about. The orientation , similar to a setup and some exposition, is followed by the complicating action their term , in which something goes awry.
This section carries the bulk of the story. Often, next is an optional evaluation, in which the speaker steps out of the narration to tell of its import a feature rare in movies, but see The Big Short, , as if to keep the listener attentive; then there is the resolution , or climax, which describes the outcome of the conflict; and finally a coda, like an epilog, which marks the close.
Moreover, it should be stressed that these oral histories are unrehearsed and spontaneously offered. Thus, people with little formal education can spontaneously generate this kind of narrative structure. In turn, this suggests a deep well of shared cultural knowledge about how stories are told. Adding generality, Longacre presented a nearly identical scheme for oral stories across cultures.
Okabayashi , for example, describes these parts. Part 1 is the ki , where the story begins and the characters are introduced and begin to interact. The opening frame typically provides an establishing shot on the location in which the story will take place, much like that in a movie prolog. Okabayashi calls this the development stage although it seems to combine with aspects of the complication or confrontation in movies. Part 3 is the ten , a dramatic and unexpected turn of events.
This seems to have the sense of a combination of the climax with a twist. Finally, Part 4 is the ketsu or conclusion. It resolves the conflict, but unlike a traditional epilog, it may leave some loose ends for a subsequent adventure in next installment of the story. Again in the medium of visual narratives, Cohn , outlined the structure of visual sequential narratives and matched five parts to other forms of storytelling. The first part is the establisher , which sets up an interaction but without any action.
Your email address will not be published. Save my name, email, and website in this browser for the next time I comment. When and how important story information is revealed. Create a timeline of your own with five stages use screenshots from the film if possible Identify the five key stages of the narrative: Act 1 Ordinary World 2.
Disruption Act 2 3. The journey or quest. The traditional way to tell a story is in a linear structure where the beginning, middle, and end happen chronologically. In film, linear narrative structures are most common. However, fragmented narratives-or nonlinear storytelling-became more popular in the later half of the 20th century thanks to the French New Wave filmmakers.
Today, nonlinear narratives have become a staple in grittier films like those directed by Quentin Tarantino, David Fincher, and Christopher Nolan. These narratives are structured in a way that allows filmmakers to tell the most effective story, rather than one that may feel stunted by a strict linear structure.
Finding the right narrative structure for your story is just as important in content marketing as it is in film. Not every story should be told with a linear structure.
Sometimes, the action happens in the middle, and that's the most logical place to begin. Structure boils down to capturing your audience's attention and holding it until the end. If you feel the beginning of your story is too slow or doesn't provide enough information to keep your reader interested, starting in medias res in the middle of things is your best bet.
Setting is a crucial part of a film's narrative. It can be argued that setting is the single most important component in an audience investing in what's playing on the screen. When the setting is right, the movie feels immersive, as if there is no barrier between the audience and film. When wrong, a setting can keep the audience from fully committing to the story. The best settings don't only pull audiences into the film-they reveal important details about the narrative. Character, plot, and thematic details can all be established within the setting in subtle and interesting ways.
Translating narrative settings in movies to content marketing may not seem obvious at first, but many of the techniques involved in crafting a setting for film can be used in your stories, too.
One of the best ways to build setting and adhere to the golden writing rule of "show, don't tell" in your story is with interactive content. Pairing multimedia elements such as video, images, animations, and GIFs, interactive content allows you to tell a more immersive story to your audience.
With many different interactive elements for your audience to engage with, it's easy to bring your setting to life. Voice is one of the strongest elements in film. It provides the most information about the characters, giving the audience a better handle on who these people are and why we should care about them. Because voice is such a fluid concept, it can be difficult to nail down.
But when you get it right, voice has the power to keep audiences entranced. It can also help establish a film's pacing and plot structure, so it's important for filmmakers to develop a voice that's dynamic to ensure the movie doesn't fall flat or feel static.
The reason digital storytelling fails is because people feel disconnected from the story.
0コメント