Frame Analysis |
ABSTRACT Frame analysis is a familiar method in the social sciences. But, what is a "frame"? Frames are the concepts we use to organize experiences. Charles Fillmore's frame semantics provides a theory of frames. Framing involves using category concepts to organize other concepts by putting them in relation to each other. mental models and mental spaces. framing Cognitive frameworks What is a frame analysis? How Do You Do Frame Analysis? The type of frame analysis I advocate is "comparative." |
Frame analysis is a familiar method in the social sciences since Bateson’s and Goffman’s delineations of it.
But what is frame analysis?
jjs: According to Carragee & Roefs (04), framing has the potential to link news texts to their production and reception. Is this a statement of the aim of frame analysis that corresponds to your goals in frame analysis?
AR: Yes, the trick is to come up with a comprehensive theory that accounts for the production, reproduction, and reception of frames.
jjs: And the theory also needs to account for “the social construction of meaning”?
AR: yes.
Marvin Minsky's conception of a frame is one of the earliest uses of the term (e.g., A Companion to Cognitive Science, 61)
A frame is a data structure representing stereotypical situations in terms of features that are always true of such situations, as well as terminal slots for features which may take on a variety of different values but must be assigned some value in a given situation. For example, in looking at a room from a given viewpoint, there will always be walls in the scene, but several options for the color of the walls. Slots will generally have default values associated with them, but these can be dislodged if alternative information is presented. The slots themselves may be filled by frames, providing for a recursive representational system. Another source of systemacticity is that frames are related to one another by transformations (e.g., the transformation of moving from one viewpoint in a rom to another). Some transformations will result in changes in the slot-fillers, while others will retain the same values. Minsky proposed that when a person encounters a situation, what the person tries to do is to match the information about the situation to a frame in a stored frame system; once a possible frame is proposed, it generates expectations that guide further search for information. If such a search produces information inconsistent with the frame, the a new proposal for a frame must be advanced. Minsky linked his notion of frame to schemata, which figured in Bartlett's (1932) account of memory, and to paradigms which figured in Thomas Kuhn's account of normal science. (61)
Although Erving Goffman's Frame Analysis is usually cited a a locus classicus for frame analysis, his conception of a frame is so broad that it is relatively useless. He writes:
I assume that definitions of a situation are built up in accordance with principles of organization which govern events—at least social ones—and our subjective involvement in them; frame is the word I use to refer to such of these basic elements as I am able to identify. That is my definition of frame. My phrase 'frame analysis" is a slogan to refer to the examination in these terms of the organization of experience. (10-11).
For him, frames are simply the concepts we use to organize experience.
For our purposes, William Croft and Alan Cruse's conception of a frame is a good starting point in understanding frame analysis.
Certain concepts 'belong together' because they are associated in experience. To use a classic example (Schank and Abelson 1977), a RESTAURNT is not merely a service institution; it has associate with it a number of concepts such as CUSTOMER, WAITER, ORDERING, EATING, BILL. These concepts ... are related to RESTAURANT by ordinary human experience. The concept of RESTAURANT is closely tied to the other concepts, and cannot be isolated from the other concepts. (Cognitive Linguistics, 7)
David Lee makes the same point using a different example:
... a good understanding of the word wicket requires a significant amount of knowledge that extends well beyond the dictionary definition. We refer to this background knowledge as the 'frame'." The frame is not in itself what is generally thought of as 'the meaning' of a word but it is nevertheless crucial to an understanding of it. ... In principle everything that a speaker knows about the world is a potential part of the frame for a particular term, even though some aspects of that knowledge base are more immediately relevant to a particular term than others (and therefore more strongly activated when the term is used)." (Lee, Cognitive Linguistics, 2004. 8)
AR suggests that in frame analysis "the trick is to come up with a comprehensive theory that accounts for the production, reproduction, and reception of frames." Fillmore's frame semantics, the source of Croft and Cruse's delineation the concept of a frame, is a foundation for just such a theory.
Outline of a theory of framing
In this context it is instructive to consider neologisms—the invention of a new word. When a word comes into the language (is considered to be worthy of an entry into a standard dictionary), what happened leading up to that event illustrates the view of "frames" in the preceding account. How, for example, did the word "blog" enter the language? Initially, a novel situation occurs which is then repeated. An advance in web technology allows persons to write collaboratively online. Many persons use this technology to enter in personal information. This is a bit unusual and in most respects unprecedented. So, its users begin to refer to this practice as "blogging" which produces "blogs." The term is bandied about and people who do not blog want to know what the term means. So, someone defines it for them. When the term is used with considerable frequency, it gets put into a dictionary with a definition such as : "a shared on-line journal where people can post diary entries about their personal experiences and hobbies; "postings on a blog are usually in chronological order.") But in describing or defining blogging, other concepts are used and they are then associated with the term "blogs," e.g., online, journal, diary, personal information, opinions, and so on.
Some of the associated terms "belong" to other sorts of experience—writing a diaries or journal, surfing online. However, if the term "blog" is mentioned before the other terms, it frames them.
Fillmore's theory of semantic frames accounts for the linguistic aspects
of a frame theory but not the social aspects. What about the person who made
up the term blog and the persons who have learned how to use it? We have
to turn to psychology to add to Fillmore's theory. So, we will look at Philip
Johnson-Laird's conception of a mental model and Giles Fauconnier's conception
of mental spaces and cognitive blending.
What is a mental model?
It is a construction that exists only virtually. Consider the mind metaphorically as a virtual terrain. The terrain has a space called “working memory” It also has other spaces called memory systems.
Let’s take the working memory as roughly equivalent to Gilles Fauconnier’s “mental space”wherein perceptual inputs trigger existing memory units. Let’s consider an input to be a sign, that is, an expression issued by someone else designed to guide us in constructing in our mental space a mental model of experience that is meaningful.
“Expressions do not mean; they are prompts for us to construct meaning by working with processes we already know. In no sense is the meaning of [an] . . . utterance “right there in the words.” When we understand an utterance, we in no sense are understanding ‘just what the words say”; the words themselves say nothing independent of the richly detailed knowledge and powerful cognitive processes we bring to bear. M Turner, 1991
The cognitive activities of perception, memory, and expression impose boundaries on experience. A precept enters the memory as an image. Images are then metamorphosed into concepts. Percepts and concepts acquire expressions by convention.
Consider children learning a language. Their perceptions are identified by gestures that point to objects. Adults then name those objects. Children then learn the conventional naming. One might consider the objects to which children most often point as figures, fore grounded by their gestures, a tree, a cat, a picture of a bird. Eventually, the figures are associated with backgrounds. Pigs in a pen. Birds in a tree. And so on.
Considered from the viewpoint of perception, the boundaries given to our experience of the world are provided by our eyes. We see through a "perceptual field" which is bounded on all sides. Objects come into the field and go out of it. The experience of perceiving is metaphorized as "taking a picture" by attending to what we see which builds boundaries around what we see and foregrounds some aspects against others in a background. Because we are able cognitively to link mental models to other mental models, we can speak of “chains”of mental models. Because we are able cognitively to embed mental models into other mental models, we can speak of “stacks”of mental models.
When we are constructing a mental model in a mental space, we give it boundaries, that is, we foreground figures against a background.
a. The background is the frame.
b. The structure of a mental space becomes meaningful when the boundaries are set conceptually.
The boundaries may go from left to right and top to bottom. If some elements of it are on top, others on the bottom, and so on, it frames the figure of a person where we expect hats to be on “top”of heads and shoes on the bottom of legs with things to his right and left. If the top is construed by convention as North, the bottom as South, the left as West, and the right as East, we may have a map of a terrain.
If we return to the metaphor of a picture frame, it shows how a frame gives
meaning to an event. for example, it allows one to picture motion as a movement
from
right
to
left, or top to bottom. It also shows ordering things, on top, on the bottom,
on the periphery, in the center. It also allows for the perception of distance.
Positions convey meaning because they convey relations. For example, we say
that a Dean is "higher" on the governing body than a department chair. Or,
Jane Fonda is "left" of Bush. Once persons or things are located in a mental
space, it serves as a mental model of the world.
How do we use these mental spaces we are now calling "frames"?
When we frame a a situation, we construct a mental space containing, and defining its proportions and positioning its components with respect to figure(s) and background.
An organizing frame provides a topology for the space it organizes; that is, it provides a set of organizing relations among the elements in the space. When two spaces share the same organizing frame, they share the corresponding topology and so can easily be put into correspondence. Establishing a cross-space mapping between inputs becomes straightforward (Fauconnier & Turner 123)
In the conventional sense a frame is a pattern that organizes its components. framing thus is to place a component in a pattern or, perhaps more accurately, to use a pattern to give meaning to one of its components. Think of taking a photograph. You are in a situation (perhaps a wedding) and you begin to take photographs. You ask the bride and groom if you can take their picture. You suggest that they stand on the lawn with the country club in the background. You instruct the brides parents to stand to the left of the bride and the grooms parents to the right of the groom. Then you decide on a different way of framing the event and suggest that the groom's parents stand next to the bride and the bride's parents next to the groom. You position the groom's mother next to the bride and the bride's father next to the groom. As you organize this scene different meanings are implied by the final photograph.
A frame is analogous to the setting of a photograph. Different settings evoke different meanings. A photo set with a church as the location adds a religious meaning to the event of a wedding. Imagine for contrast, a photo of a bride in a bathing suit that suggests a bridal gown and a groom with a bow tie around his neck and creased swimming shorts to suggest a tuxedo with the ocean as a setting. Religion is not the first association one would have with such a photograph.
A frame is a mental schema, an identifiable unit in memory, a meaning structure (conceptual or narrational) that includes other meaning structures. In this sense, a mansion is a frame containing the concept, house. So also is a motif as in the femme fatale motif. Hence framing means to contextualize a meaningful unit of language by locating it within a "larger" unit. A house can be categorized as a mansion or as a hut. Categorization seems to be the underlying cognitive activity. See Langacker, Foundations of Cognitive Grammar on "abstraction" [3.3.3] and on "transformation" [3.4]
In scanning a stream of conscious awareness of our environment characterizes our experience. We can attend to specific elements in our mental model of it, focus our perspective, and structure it. We can compare these mental models of experience. We can link them to other similar models and we can transform them into structures other than the ones given in experience through imagination. We can create metaphorically more “abstract”models and include more specific models in them. In addition, we can store these models in our memories and retrieve them when needed to guide us in our experiences of our world.
These mental models derived from our stream of consciousness are “partial representations”of experiences. As John Dewey long ago pointed out, our experience comes to us in a stream of consciousness out of which we select “an”experience upon which to focus and which we can represent as a “model”of the given experience. “Things are experienced but not in such a way as they are composed into an experience,”he wrote Moreover, we can express these models of our experiences through signifying systems as long as others share our expressions through convention.
We use categories to organize our experiences of the world. We tend to structure them by using overarching schemes common to our cultures—time and space, calendars and geography, life spans and dwellings. These larger structuring structures are cognitive frameworks that organize our frames. At the most general level (worldviews, historical periods, geography) are a shared by members of particular cultures.
At the moist general level, it is a description of the organization structure of a representation of experience, of which the most common are verbal. Frame analysis is the most rudimentary form of discourse analysis.
Going back to the photography analogy, we can analyze the outcome of taking a picture, the photograph with respect to its structure, with respect to how its components are related to each other. Placing the bride's parents next to the groom is a different structuring than placing them next to the bride. In the first structure (B'sP + G + B + G'sP) the meaning that can be derived from this positioning is that by getting married you not only form a bond with the person you marry but also with that person's parents. The second structure (G'sP + G + B + B'sP) suggests a union of two families. If the photograph were of the two sets of parents (G'sP + B'sP), it suggests that they have formed a new relationship based on the marriage of their children.
In these remarks, we are analyzing the relations among the components in the photograph. If we introduced a new component to the second photograph (G'sP + G + B + B'sP), for example bottles of beer, we could frame the relationship between the families somewhat differently. For example, [(G'sP + b) + (G + b) + B + B'sP). This suggests that the Groom and his family are drinkers but that the bride and her family are not.
This last example points to the problem of frame analysis. The circumstance that the bride and her parents do not holding beers does not necessary mean that they do not drink. Yet, the photograph does relate the Groom's family to drinking beer and not the bride's family. Was this intentional or unintentional? To decide this issue, one has to go "beyond" the photograph.
The photograph has a structure independent of the photographer's intention. But, can be predict that someone looking at the photograph will draw the inference that the bride's family are teetotalers? We can't.
What's the solution?
Basically comparative analysis.
If,
for example, you had all of the wedding photographs taken by everyone there
and you noticed that in none of them
did the Bride's family have an alcoholic drink, the odds that they didn't
drink would go up significantly. And then, if some of the photographs were
of the wedding toast and everyone except the bride and her family had champagne
glasses in their hands, the odds go up every higher. Let's then say that
you had access to the family photo albums and again no evidence of drinking
in any of them even en though in the pictures other persons were clearly
drinking. The odds go up again.
[The Bush/Gore photos]
Returning to discourse, for the purposes of illustration I offer a brief illustration of comparative frame analysis.
text #A:
A1: Jack is something of a ladies-man. I feel sorry for Jill. If she knew about his goings-on, she would be crushed. (John)
A2: Jill is a very beautiful woman. It's a pity she's married to Jack. How she ever fell in love with that jerk, is beyond me. (John)
A3: Harry told me a funny story about Jack. At Ellen's party last week, he tried to pick up Joan. She told him to "get real." (John)
A4: Jack is a womanizer. I've wanted to tell Jill many, many times but I thought it would hurt her too much. (Joan)
text #B
Michael Vick, the glamorous quarterback of the Falcons, apparently did some unglamorous things in his spare time.
Michael Vick, head shaven, smartly dressed, appeared in court to face charges of financing dog-fighting.
Michael Vick, an African American, is not from a mainstream culture. For him and his brother, dog fighting is a sport. We kill deer for sport.
text #C: Historic Building to be Torn Down
C1. The Kramer building on Wabash will be torn down later this month to make way for the construction of an eleven story apartment residence with shops and restaurants on the ground floor. (The Financial Times)
C2: The historic Kramer building, constructed by Louis Sullivan in 1893 will be destroyed before concerned citizens have an opportunity to organize a protest against the realtors who plan to erect an apartment on the site. (Historical Architecture)
C3: The oldest remaining structure on Wabash, the Kramer building, will soon be replaced by luxury apartments with attractive shops and new restaurants on the lower floors of the eleven story high rise. (Sun Times)
C4: Thanks to the avarice of our city managers, next month one of the few remaining Sullivan buildings, which is one of his most elegant structures, will be lost, never again to be seen by visitors to our city. (StreetWise)
C5: Too costly to be renovated and becoming increasingly decaying structure, the well-known Sullivan building will be officially condemned and then torn down. (Chicago Tribune)
text #D
D1: If Greise were the Bears quarterback, they would be in a better position to make the playoffs.
D2: Sportscasters around the country believe that Rex Grossman is the biggest obstacle the Bears face in returning to the super bowl.
D3: You can't say that Grossman lost the Bear's opener, but you can't say we won it either.
D4: Ron Rivera told his linebacker that Grossman can be rattled and Phillips did certainly rattle him.
According to Carragee & Roefs (04), framing has the potential to link news texts to their production and reception. I will take this to be the principle aim of this outline.
PRENOTES: Linking a news text to its producing agent in order to infer the agent’s “intention” would be on several counts virtually impossible. In the first place, any news text cannot be said to have a single agent as its producer. Even in instances where reporters have signed particular texts, they must be considered at best as a spokespersons for the company networks that make up the production cycle. As such, their writings are subject to a variety of constraints which are not usually made public. Style, format, tone, length, and other generic constraints apply. Legal constraints apply. Audience constraints apply (the expectation of a particular point of view associated with the new media). Nor is it invariant that the utterances of a particular agent were composed by that agent. For example, a TV reporter works from a script which is co-authored. A reporter’s text might be edited by the section editor. And so on.
[This is acknowledged by Carragee & Roefs as well as by Gamson & Modigliani. However, Gamson speaks of “the intent of the sender” (158) although he allows that the sender is a composite of several actual senders.]
Second, with respect to reception, outside of interviews with actual readers or viewers, the ways in which they have construed a particular news text cannot be accurately inferred from the words they heard or read.
I will take as my sample text the following passage used as an example of “the progress frame” by Gamson and Modigliani (1989)
There has always been resistance to technological progress by nervous Nellies who see only problems and ignore the benefits. Resistance to nuclear power is the latest version of this irrational fear of progress and change, the expression of modern pastoralists and nuclear Luddites. The failure to develop nuclear power will retard our economic growth and make us renege on our obligation to the poor and to the future generations.
Various readers can construe this text in ways that have little to do with the lexical and syntactic form of the text. For example, I read this text as blatantly polemical and find the subject position assigned to me by the syntax—namely that of a “nervous Nelly,” “pastoralist,” and “Luddite” so crudely proffered that it reads as a satire on itself, something I might expect watching John Stewart.
[Gamson speaks about a “preferred reading” that corresponds to the “intent of the sender” (158). This seems problematic. ]
The type of frame analysis I advocate is comparative.
Beginning with a “situation,” (an event) the analyst examines texts commenting on it. It is a comparative method. It examines commenting texts for story structures in which the event is located looking for ones that are frequently repeated. Then it describes the narrative functions within the basic story structure. It then looks for repeated discourse themes associated with the story, noting which agents are their sources. Then it looks again for embedded stories in the “frame” story or “counter” stories identified by reverse outcomes. In an empirical study, the analyst would then ask a sample population who were exposed to the texts to “explain” the initial situation or to “tell its story.” Their texts would be analyzed for matches to the framing story. One would also analyze statements from “sponsoring” agencies for matches.
Since frames underlying all discourse structures, a comparative analysis identifies the structures and notes in particular how they begin. The beginning frame is usually the one in which the subsequent frames are included and positioned.
A comparison of discourses will reveal how each of them instructs their listeners or readers to construe them. The instructuring provides the "hypothesis" that needs to be investigated empirically through ethnographic techniques.
jjs
Notes:
. [Bateson
understands a frame as a “meta-communicative”, communications
about the communication. (Bateson, 1972a). He understands
a frame as the communication from “a higher level
of consciousness”to a lower level to indicate how to contextualize the
ongoing event. From the point of view of CL, the unconscious is the memory
system that supplies framing by locating expressions in networks that can be
verbalized as semantic domains or scripts. Bateson is in a tradition that
employs theories of formal logic and mathematics to natural language issues. His
approach is unusual because his perspective on the problem of framing is psychoanalytic.] ... 
. [The
term has a vexed history of definitions. See Deborah Tannen’s “What’s
in a Frame,”Framing
in Discourse, 14-56.] ...
. [A
frame is a metaphor when used to describe cognition. It suggests that we draw
boundaries around our experiences in order to create networks of
meaning. The metaphor of a frame is profoundly spatial. If we accept Fauconnier's
theory, a frame structures a "mental space." (Fauconnier, 1994)
. Mental spaces are “separate domains of referential structure.”(Fauconnier,
1994) Mental spaces are derived from our perceptions. Frames
presuppose references to experiences. Basic mental spaces are representations
of the world as we perceive it. These mental spaces are used metaphorically
to create more abstract meanings (meanings that are not directly derived
from perception) [Lakoff's view of Metaphors We Live By] Mental spaces are
stored in memory systems. A memory system is a network of related mental
spaces. Mental spaces are verbalized in our language systems. ] ...
. [
Our language system has primary types of relations that allow us to link frames
(mental
spaces) Langacker identifies basic ones: motion, order, distance.] ...
. [Frames
also are the conditions for comparisons. ] ...
← previous | next →
last revised:
June 13, 2007
Send comments to jjs.
copyright © jjs, 2007