SOCIAL COGNITIVE REPRESENTATIONS OF VIOLENCE IN URBAN BRAZIL: VIOLENCE AS AN UNCONTROLLED FORCE

 ABSTRACT: The present paper reports on partial findings of a major research project entitled Interdisciplinary project on social cognitive representations in the conceptualization of violence in urban Brazilian centers . 1 It aims at providing some insight into an embodied and social-culturally situated concept of representation – a social cognitive representation, as I understand it, by presenting both a theoretic-philosophical basis for this kind of concept as well as some preliminary findings of an analysis of focus groups talks about urban violence. The methodology is focused on discursive and cognitive aspects which emerge from and influence people’s verbal interactions. The qualitative analysis performed suggests, among others not discussed here, the emergence of two metaphors in participants’ talks: VIOLENCE IS A PRODUCT MANUFACTURED BY THE MEDIA and VIOLENCE IS A SPREADING CONTAGIOUS DISEASE. This appears to be due to the Media’s sensationalist trivialization of violence which results in violence being thought of as present anywhere and everywhere like an UNCONTROLLED FORCE. The trivialization of violence by the Media contributes to negative attitudes and feelings of lack of control, distrust, prejudice and


INTRODUCTION
The conceptualization of violence, even when a specific form of the phenomenon, such as, "domestic violence" or "urban violence" is under analysis, is a complex task. This is true because embodied, social-culturally emergent concepts do not display sets of well defined mutually exclusive, essential features as the classical theory of categorization would make one believe (LAKOFF, 1987). Thus, rather than departing from aprioristic tentative definitions of what the concept of URBAN VIOLENCE might entail, and then trying to check these in comparison with our data, it is our intention in this paper, to focus on conceptualizations which emerge in representations of the phenomenon as evidenced in verbal interactions of individuals who have to live with the threat of urban violence on a daily basis in Fortaleza, a fast growing city, capital of the state of Ceará, in northeast of Brazil.
The paper thus, focus in discussing, under the light of Lakoff's theoretical proposal known as Idealized Cognitive Model (ICM), the nature of what I term social cognitive representations, that is, embodied and social-culturally situated mental evocations of the concept under analysis. In order to achieve this aim, in the next section I present a brief description of the classical concept of mental representation which has been greatly influenced by Descartes rationalistic thesis, then Lakoff's (1987) ICM theory is synthesized and a claim is put forward as regards the pertinence of such a proposal as a basis for the concept of a situated mental representation founded both in the capacities and limitations afforded by the body's nature as well as in the ecological and social-cultural environment in which the individual is situated. This discussion is, then, followed by a description of the methodological procedures adopted for the study, the data analysis and results and, finally, the concluding comments and references.

ICM THEORY AS SUPPORT FOR AN EMBODIED, SOCIAL COGNITIVE SITUATED REPRESENTATION
ICM theory provides the basis for a reformulated view of mental representation which opposes the classical thought firmly engrained of such representation as an idealized and decontextualized mental construct. Before the argument for an embodied, situated cognitive representation based on Lakoff's (1987) proposal is put forward, a synthesis of the classical view of representation, with emphasis in Rene Descartes ideas, is outlined in the following paragraphs.
Descartes thesis of a rational mind has had throughout the centuries, a very strong impact on people's general beliefs as regards the ways humans represent reality internally. According to the philosopher, representations engendered internally were the result of a transcendent mind, one which was in no way dependent or limited by the body. In fact, Descartes considered man a duality composed of body and mind. While the latter was an ethereal substance infused in the physical body, the former was material, sinful and limiting, simply, the instrument used to keep contact with the world. (DESCARTES, 1984). According to this view "the real knowledge of external things seems to belong only to the mind instead of the mind and the body" (RYLE, 1949, p.305, apud PERRY;BRATMAN, 1993).
The dualistic view of man was largely adopted by western philosophy, and also influenced Cognitive Science greatly since its beginnings in the 50's. The idea advanced by this field of research was that man was able to (re)present the external world internally and that such representations would be engendered in the mind by the computation of symbols that corresponded to external realities. These symbols manipulated and processed internally by the mind were, however, separated and strange to the external reality, therefore, not part of it. Symbolic representations failed therefore in a crucial aspect. Their separatedness from the reality they were supposedly able to (re)present conferred such representations a special nature which distanced them from the world they mirrored. However, the problem was "how could abstract symbols which the mind was able to manipulate by algorithmic rules, represent meaning"? How were human intentions, values, beliefs, desires, represented by abstract symbols engendered in a mental syntax? This problem is well highlighted by Teixeira (2004), when he asserts that, [...] the intentionality problem is insurmountable for symbolic computation. More than that, it sets the limits of the classical theory of representation adopted by cognitive science or by cognitivism it its beginnings. Symbolic manipulation, no matter how sophisticated it may be, is not able to make symbols refer to or come to be "about something in the world" (TEIXEIRA, 2004, p.47-48) The classical theory of representation fails to provide a solution to the problem of intentionality since, as clearly highlighted in the citation above, the mere filing of information in the form of symbols and their manipulation in memory do not contain what is extra-mental and therefore cannot establish "the relationship between representation and its referent in the world" (TEIXEIRA, 2004, p.48). How can such an impasse be resolved? I believe that ICM theory can provide some profitable insight into an alternative form of representation, one that emerges from the dynamic interactions of a cognoscitive being in the world and which considers such a being not as a duality but The citation above nicely synthesizes the argument put forward in this paper. That which sees ICMs as a valid and adequate explanation for a concept of representation as an emergent cognitive phenomenon which is made possible from the cognoscitive beings embodied situatedness in their ecological, social and cultural world. Whatever nature a mental evocation might assume, it occurs not as a strange entity in relation to the world, but, emerges from a distributed (or shared) kind of cognition which results of what our bodies afford us in the environment in which we live.
Evidence in favor of such a view of representation will be presented in the section where the corpus analysis and results are discussed, but, first I provide in the section below, some insights into the methodological procedures followed in the research.

CORPUS, PARTICIPANTS AND SOME METHODOLOGICAL DECISIONS
The research corpus, which is partially analyzed here, is composed by verbal interactions produced by 11 voluntary postgraduate students, aged 16 to 43 years old from two private colleges and one public university in Fortaleza, Ceará, Brazil. They interacted in two focus groups around the topic of urban violence. Among the questions they discussed were "When you hear the expression 'urban violence', what's the first thing that comes to your mind? "Who do you talk to about urban violence?" "In your opinion, what have been the effects of urban violence on people?" The first group was moderated by a research assistant and the second group, by myself. The idea was to put the questions to the groups and then let the participants interact around the topic by themselves as much as possible. Interventions by the moderators were few and had the sole purpose of posing the questions and bringing the discussion back into focus when necessary.
The questions were translated from English into Portuguese. They were adapted from a list of questions devised by Lynne Cameron in order to research people's perceptions of terrorist attacks and their after effects, known as Perception and Communication of Terrorist Risk (PCTR). 2 Since the research in Brazil had urban violence as its object, besides being translated into Portuguese, the questions had to be adapted in order to make them proper to investigate urban violence -the phenomenon analyzed in Brazil.
The methodological decision to use data gathered from verbal interactions was guided by the theoretical option I take of conceiving of mental evocations which participants might construe about urban violence as situated embodied representations which emerge from discursive activity, but, which are grounded in shared values, ideas and beliefs pertaining to the social-cultural context of the participants.
Attention, thus, centered on the language produced by participants as they interacted verbally in providing responses to a moderator's questions about urban violence. Special attention was paid to the figurative language produced and the types of social cognitive shared attitudes, values and beliefs which appear to function as grounds for such language.

METHOD AND PROCEDURE FOR DATA COLLECTION AND ANALYSIS
The methodology adopted here is both discursive and cognitive. The decision to follow a mixed procedure was motivated by the fact that I believe that in order to grasp the scope of cognitive representations more fully it is necessary to view cognitive and linguistic activity as happening both as a bottom up (discourse-cognition directed) as well as a top down (cognition-discourse directed) process.
The discursive perspective taken for the data analysis is the one known as metaphor-led discourse analysis 3 (CAMERON ET AL., 2009; CAMERON; MASLEN., 2 The PCTR question list was devised as part of a UK project funded by the European Social Research Council (ESRC) under the New Security Challenges research program which took place from 2005 to 2007. The project main concern was to investigate the ways in which people perceive the risk of terrorism and the consequences for official risk communication. Since the questions in the UK project were geared towards terrorism, the questions were adapted or changed in order to make them suitable for investigation in Brazil. 3 Metaphor-led discourse analysis is a term used by Cameron and colleagues to promote the idea that metaphor analysis has to be approached from a discourse perspective. Discourse here is to be understood as what happens when people engage in talk and dialogically make meaning emerge from such interactions. According to this perspective, metaphor is viewed as a dynamic multi-faceted phenomenon. Under such a view, metaphors, like discourse, are understood as "linguistic and cognitive processes, 2010). I view such a procedure as adequate for the analysis to be reported in the next section since metaphors which emerge from linguistic interactions are indicative of what are cognitively shared by the group.
The association of intertwined bottom-up and top-down processes in tentatively deriving the kinds of social cognitive representations participants have of violence is in harmony with a situated view of adaptive behaviors which instead of resulting from a duplication of reality in terms of formal propositional internalized representations are evidence of virtual activations of motor and perceptual procedures which influence and are influenced by the social-cultural and ecological environment. (GIBBS, 2006;GIBBS;CAMERON, 2008). It is, thus, assumed that the figurative language used in the interactions is to some extent anchored in socially shared beliefs engrained in ICMs which are part of their reality.
Three related research questions are the focus of the analysis which is reported in the next section: (1) What are the key themes around urban violence that emerge from the Brazilian focus groups discussions and how to they express these by means of language? (2) What metaphors emerge in their talk? (3) What kind of cognitive models seem to motivate the language produced by focus groups participants?
In order to answer these questions participants' verbal interactions were video recorded. Their talk was then transcribed and fed into Atlas.ti 4 in order to search for, discourse topics (DTs), metaphor vehicles 5 (MetVs), and metonymies (Metons) and organize them into families.
The methodological steps followed for the organization of data in Atlas.ti and discursive analyses were as follows: (1) Participants' utterances during the focus group discussions were broken into simplified intonation units (IUs), with endings and pauses. In order to do this, we followed Cameron et al.'s (2009) procedures. According to the authors this representation of the discourse strings produced during online interactions, try to capture "something of the temporal dynamics iconically through the use of intonation units and layout" (CAMERON ET AL. 2009: 14). For the present research purposes we marked the pauses by inserting the following punctuation mark (.) to the beginning of the next string of talk and assigned one dot (.) to each second the participant took before expressing her next idea 6 . Still in line with Cameron et al.'s (2009) methodological procedures, when the focus group data included instances of quasi-reported speech, which happens when a speaker adopts the voice of some other person or organization; these utterances were enclosed in <Q ... Q> brackets. The symbol <X...X> represents a stretch of speech which was indecipherable to the transcriber. flows or movements" which are part of complex, dynamic systems (LARSEN-FREEMAN; CAMERON, 2008; CAMERON ET AL., 2009). 4 Atlas.ti is a qualitative analysis software designed for the organizing and coding of data into analysis categories. More information about the software can be found at www.atlas.ti.com 5 According to Cameron et al.'s (2009) methodology here adopted, metaphor vehicles (MetVs) refer to discourse words or phrases with incongruous meanings in relation to their most basic (concrete, experiential) meanings. The procedure is based on MIP (method for identifying metaphorically-used words in discourse devised originally by the Praglejaz Group (2007). 6 For a full explanation on the procedure followed in the identification and marking of the IUs in Metaphor-Led Discourse Analysis, we refer the reader to Cameron et al. (2009), listed in the reference section of the present paper. Revista da Anpoll nº 37, p. 293-316, Florianópolis, Jul./Dez. 2014 (2) Since the present research is developed as s subproject of a larger one entitled Living with uncertainty: Metaphor and the dynamics of empathy in discourse, coordinated by Professor Lynne Cameron, from the Open Universit in the United Kingdom, all the discourse produced by the Brazilian participants was translated into English by a proficient Brazilian Portuguese speaker of English. This was necessary in order to make possible the simultaneous cooperation between the Brazilian and the English research teams as well as for research publication purposes. The English translations appeared in blocks after stretches of data transcribed in Portuguese. (3) Each line of the Portuguese data, corresponding to an individual IU was numbered in order to facilitate the location of important information and as a requirement for data uploading in Atlas.ti.
The sheets of transcribed discourse were identified with the following information: data of focus group interaction (i.e. number of people present, age range and academic affiliation and date of the interaction).
Discourse produced by the Brazilian participants was fed into the Atlas.ti software. DTs, MetVs and Metons were identified and coded so that systematic metaphors (SysMets) could be arrived at by the analyses of the language produced by participants. (6) Once uploaded in Atlas.ti, all transcriptions were completely read and possible key discourse themes were noted. The transcribed data comprised three individual documents identified as P1 and P2 -data gathered by the interaction of the FA7 Focus Group (FG1) and P3, for the data gathered from the FIC/UFC Focus Group (FG2). The overall reading of the documents and identification of key discourse topics (DTs) allowed for and understanding of the discourse event structure. The next step was to break the whole discourse event into sections and describe what was happening in each one of them.
A thematic analysis was then performed and MetVs and Metons groupings were organized in order to analyze the figurative language which emerged during the verbal interactions. This was necessary so that systematic metaphors could be proposed and analyzed.
In order to prioritize accuracy and since the methodological procedures and data analyses are carried out as teamwork performed in connection to the various subprojects which are encompassed under the major researches in Brazil and the UK (see note 1). After the steps above were performed, the preliminarily coded transcripts were submitted to reliability checks, as follows: • 10% of the entire data was checked; • The process was carried out (blind) by at least 1 other researcher • The results were compared and possible disagreements were discussed; • Once an agreement was reached on possible differences, decisions were noted for the rest of the work.

Linguistic and cognitive aspects considered in the analysis
As regards the analysis of cognitive aspects related to the talk produced, attention was paid to participants' conceptualizations, their feelings, ideas and beliefs as these emerge in their talk about urban violence. It is felt that these provide important insight into the kinds of cognitive models participants let emerge in their talk. In order to provide evidence of the interconnection between language (discourse) and cognition, participants' verbal interactions are linked to propositional and imagetic schemas which are part of the socio-cognitive representation of urban violence which emerges from the data.
ICM theory (LAKOFF, 1987) and Johnson's (1987) schema typology provide the theoretical and methodological basis for the analysis of cognitive aspects in the data.
Therefore, concomitant to the methodological steps listed above for data organization and analysis, cognitive schemas pertaining to socially and culturally shared ICMs, which appear to be part of social cognitive representations and which might have influenced participants linguistic interactions are identified and excerpts of the talk produced by the participants are used to validate these. Finally, a social cognitive representation of urban violence which appears to be linked to the SysMets analyzed and the cognitive schemas identified is proposed.

Data analysis
The reading of the two focus groups verbal interactions, made possible the identification of 11 DTs. These were: change; criminals; family; government; location; media/TV; others; police, response to violence; types of emotion; violence.
Linked to these DTs were 06 Meton families and a much larger number of MetVs which were organized into 36 family groups. The Meton families included labels such as, acting stories, body, change, inside/outside, people/government, with words or expressions such as, disguise, "own hands", traffic light, "go out of the house", "public security" etc. The MetVs were grouped into families like, body, change, connect/separate, constraint/control, disease, and included words and expressions such as, castration, violate, "on my side", "house arrest", suffer etc.
The recurrence and interconnectedness of some of these MetVs within and across DTs, along with the salient presence of the INSIDE/OUTSIDE metonymy, in special, made possible the identification of 05 Due to space limitations, but, also because of their salience and interconnectedness, the first two SysMets above will be the focus of the analysis here reported.
Below, I analyze stretches of the verbal interactions which serve as basis for the SysMets proposed. Whenever possible, based on the language which emerge in the interactions, figurative language as signed by MetVs vehicles are linked to ICMs which might be motivating the social cognitive representations participants have of violence in urban Brazil.
The participants' verbal productions which are analyzed appear below in tables in both Portuguese and English. One stretch of talk which enables the proposal of a SysMet is presented and then additional evidence for the metaphor present in other participants` talks is included in the comments following the presentation of the table to illustrate the presence of the metaphor proposed. ...that the robber feels good for being in the media ISN'T IT, Table 01, above contains excerpts of the discourse produced by Ana, the first participant in the first focus group (FG1) to engage in the verbal interaction in response to the moderator's question as to what comes to mind they hear the expression "urban violence". Her talk along with excerpts from other participants which will be commented in due course indicates that TV/Media has played a negative role in propagating violence and violent behavior.
Here, it is interesting to notice that in expressing her opinion as regards what comes to mind when she hears the expression "urban violence", the participant, by the metaphorical use of "see", instantiates the metaphor UNDERSTANDING IS SEEING, signed by Grady (1997) as a primary metaphor. It is also worth highlighting that she uses MetVs such as "instrument" and "propagation" and highlights that the TV is in every home teaching people how to make violence related implements and act violently. Ana says that TV plays a role in teaching people what is bad "you learn on TV how to make an assault …how to make a bomb" (P1, p.37-38). She states that aside from teaching people how to engage in violence, TV has played an important role in the inversion of social moral values. For instance on P1, p.62-66 (see Appendix 1), she says 7 The information presented in parenthesis refers to the Atlas file, the transcription lines and the participants pseudonyms. 8 A more native-like translation for the expression "transmite" used in Portuguese, would be "broadcasts'' instead of ''transmits'' as translated. However, in order to allow for the identification of MetVs and Metons used in Portuguese, translations are kept throughout the transcriptions as close to the lexical choices and structures used by the participants as possible.
"things that, according to your moral values, were wrong, when you see them repeatedly on TV, they come to be considered routine. It becomes normal if you rob or if you traffic drugs". Ana's conceptualization of "urban violence" is deeply affected by the negative effects TV has had in promoting violence. The prevailing influence of the media in trivializing violence and bombarding the public with it has not only taught people violence but also caused an inversion of values. Violent or outlaw behaviors that, years ago, were considered immoral now become routine and are viewed as normal acts. The negative impact of Media/TV in producing and publicizing violence is reinforced by other participants' talk. For instance, Celso, another participant of FG1, talks about the media as fabricating violence. He says "just adding more to the comment Ana made …the media issue I find that it fabricates ..violence […] ..when we go out what ..comes to mind ..it is fear, ..it is persecution ..right, so it is the values inversion as Ana explained (P1, p.84-91) (Appendix 2). Mateus, a member of FG2 pursues the same line of thought about the media by saying, how the media saturates people`s lives with it. "…so they're things that are spread from the time you switch on the TV in the morning ..to the last program you watch be it a soap opera or ..whatever you're going to watch …we eat, drink and live violence. From my point of view, it is even utopic, the fact is that violence is not as aggravated ..as we think it is ..it is just more disseminated (broadcasted) than realized (really taking place) (P3, p.121-135) (See Appendix 3) Corroborating with Mateus, Lívia, another participant of FG2, highlights that "so, then, violence ..it is present and I see that it is the media that makes a point for it to be each time more present in everybody's lives, (P3, p.206-210). She adds, in sequence, that the media has blown violence out of proportion (P3, p.212). She uses the Portuguese verb "magnificar" (to magnify), that is, the media has made violence be seen, felt and thought about as something pervasive, present and big in everybody's lives. As a result, she concludes her argument about violence pervasiveness and trivialization by the media by highlighting the violence programs in Ceará TV. She says on (P3, p.217-242) that "We see the time the TV programs on violence are on, […] They're on right at lunchtime, ..it's the time you come from work or the time you turn the TV on ..and people don't care if they're eating and watching the programs that ..are not allowed to show clear images, so they make them blurt but they end up on TV anyway. For the population, it is a little ..it's trivial now, because there's a body there lying on the ground and there are a bunch of boys ..waving good-bye, jumping, ..it's like it became (turned) something normal for them, people don't ..get scared with that anymore". (Appendix 4) The stretches of verbal interactions presented above, reinforce Lívia's viewpoint of Media/TV as an instrument in the propagation, cheapening and trivialization of violence. It does not report in a neutral, balanced way about violent incidents, it glamorizes violence and blows it out of proportion. Media invades people's lives with violence, creating in them a sense of "normality" for violence. Such reflections on the role of Media/TV on spreading and cheapening violence allows for the proposal about the presence of the SysMet: VIOLENCE IS A PRODUCT MANUFACTURED BY THE MEDIA in the two FG talks.
As regards the cognitive models which appear to influence participants` talks, the FRAME schema which is a type of propositional sub-model and the FORCE schema can be identified in some of the participants' discourses. For instance, often participants created scenarios to contextualize their comments and support their opinions and arguments. In this regard, socially shared stories are present in their discourses. These stories provide the frame to support the point they want to make. To illustrate: when Igor from FG2 actively engages in the verbal interaction with Mateus and Lívia, he brings out Isabella Nardoni's story "by this discourse, we can see that ..the Media needs these facts ..it needs this violence, such as this case of the girl Isabella, which was… At this point he is completed by Lívia who intervenes to complete Igor's talk and adds two other of these shared stories by saying: "a national ...a national case and not for a short time, ..until today people talk about Isabella, ..Eloá, and now there's this case about Elisa 9 , which is going to last for a long time" (P3, p.190-205).

(Appendix 5)
The frame sub-model is illustrated with the first of the shared salient stories brought into the discussion in order to argument in favor of how Media/TV manufactures and spreads violence. Although the narrative refers to a case involving domestic violence, it is used by the FG participants in order to illustrate how Media enlarges and hammers violence into people's lives. Here, it also serves to illustrate the cognitive organization of the frame involved in typical acts of violence. The above characterization of the frame sub-model, allows for the generalization concerning a more general frame for violence, be it urban or of another type. It can be ascertained that cognitively this frame will involve at least two partiesan aggressor and its victim (the agent and the patient, in the model's jargon). It will also be perpetrated by the use of some instrument. In the case of urban violence, evidence from Gondim's (2012) research points to firearms as the most prototypical instrument used in this kind of violence. The frame also makes provision for the violence outcome -its effect (the target or goal) which is to cause some damage in the victim (patient).

ISABELLA'S CASE FRAME SUB-MODEL HAS THE FOLLOWING STRUCTURE
Another cognitive sub-model which appears to be present in the participants' talk is the FORCE schema which can be conceived imagetically as a spreading uncontrolled entity with powerful invading force. This becomes evident in the participants' talks; for example, in all the excerpts highlighted before, one idea is pervasive -that of violence as being shoveled down into people's privacy. "TV is in everybody's houses" (Ana, P1: 35); "so then, violence ..it is present I see that it is the media that makes a point for it to be each time more present in everybody's lives, (Lívia,P3,. Ana Cristina Pelosi, Social cognitive representations of violence in urban Brazil: violence… Along these lines, Vania, a FG2 member in DT: Response to violence, speaks of violence effects as a force which builds up in people's body in the following terms "when something as a robbery or a car crash happens, we use this fact as a ..a ..a place where we can let it overflow, so ..you end up being more violent than the violence you suffered (received), it accumulated so much that you didn't know how to work on that and you don't accept it" (P3, p.1942-1955). Finally, Renato, another participant of FG2, speaks of violence effects on the ordinary person in very vivid terms. The person becomes a victim of this uncontrolled force and as a result "you cannot let this revolt overflow you accept ..you swallow all that revolt, and it is kept (imprisoned) in you and you end up letting it off (overflow) in a different way in someone that has nothing to do with it. (P3, p.2124-2130). This participant's talk brings out nicely both a supposedly primary metaphor ACCEPTING IS SWALLOWING (Grady, 1997) and another image schema, which is also shared by Vania -the CONTAINER schema, which is part of the conceptual metaphor ANGER IS A FLUID KEPT UNDER PRESSURE IN A CONTAINER. According to Renato AND Vania, frustration brought about by the general feeling of insecurity which spreading violence brings is conceptualized as pressurized revolt which might come to the point of overflowing and be let out indiscriminately. (Appendix 6, 7).
Another interesting fact to notice about the presence of cognitive models in participants talks is the fact that circulating shared stories, such as the ones about Isabella's and Eloá's violent deaths emerge in their verbal interaction as cognitive metonymical sub-models which symbolize a more encompassing prevailing state of violence portrayed by the media.
Let's now, move on to the second SysMet to be analyzed here -VIOLENCE IS A SPREADING CONTAGIOUS DISEASE. ...people will be affected The SysMet above is trigged by the first metaphor in which violence is conceptualized as a product manufactured by the media, as we shall highlight in the course of the analysis. It is proposed with basis on the recurrence and relatedness of some MetVs used by the Focal Groups participants.
Susan's opinion in table 2, under DTs: People/Others and Response do Violence, sets the scene for the emergence of this metaphor. She uses the word "affected" to describe the media's effect in disseminating violence as a prevailing widespread product which, as a contagious agent, affects everybody. In this regard, Celso's words next (DT: Types of Emotion), help to reinforce this argument. He says: "And I add more …we have a prejudiced look …we create this prejudiced look about everyone …on account of violence …it used to exist because we are prejudicial …generally speaking we're full of prejudice …now about this clearer issue …the issue of violence which is in evidence …your view (look) is prejudicial …it is always prejudicial …if someone comes close …you already stay away from the person (P1, p.455-466). He adds a withdrawing gesture with his hands and body to his words. Celso is thus, speaking of the prejudicial attitude which becomes part of people's behaviors as a kind of spreading disease. Like a contagious disease, the distrust brought about by the prevalent state of violence which to some extent is caused by Media's trivialization of it creates an attitude of general prejudice. People start viewing the other as potential violence agents and as if they are carriers of a contagious disease, do not want to get close. They try to avoid contact and withdraw. It is worth noting Celso's next words. He speaks of an incident which happened to him at a bus stop. In part, this is what he says: "I have gone through several situations of me getting closer to a lady at the bus stop …and she was very calm …I got closer …and I had a book […] …the whole outfit, …I was dressed as a student or as a teacher …a serious person …but then when I got closer, this lady made a gesture (as if she was trying to hide her handbag) and that made me feel bad …you know why I looked at her this way …I almost went to her and said: for God's sake, you lady looked at me, you grabbed your purse and made this gesture …Then I kept thinking …my gosh …we're so scared …by this word violence […]" (P1, p.467-484). Distrust, prejudice and fear prevail as a result of violence spreading like a contagious disease. At the same time, hurt is felt by those who are the target of prejudice, like Celso, in this instance. (Appendix 8) Again the FORCE schema which underlies the conceptualization of violence as an uncontrolled spreading force emerges in the participant's talk as a cognitive submodel of urban violence. Although the focus is on violence as disease, it forcefully (with the intensity of an uncontrolled force) affects and contaminates people. In this regard, the body language Celso uses in order to make his narratives on P1, p.455-466 and P4, p.467-484, more vivid serves to illustrate this. When he speaks on line 466 of people withdrawing from him, he puts both hands in front of his body and distances them from him (a repelling gesture), just as if trying to block a forceful agent (a force) coming in his direction. In addition, when he speaks of the gesture the lady makes with her handbag, again, it suggests the emergence of the FORCE schema. The lady with her handbag, as Celso portrays her, with a body gesture, distances herself from him as she tries to repel and at the same time protect herself from Celso, who, for the lady, is a potential violence agent (line 476).
The interrelation between SystMet1: VIOLENCE IS A PRODUCT MANUFACTURED BY THE MEDIA and SysMet2: VIOLENCE IS A CONTAGIOUS SPREADING DISEASE is evident in the following excerpts produced by Mateus (P3, p.121-123) which was already mentioned as part of the analysis for SysMet1. "…so they're things that are spread from the time you switch on the TV in the morning ..to the last program you watch" and "the fact is that violence is not as aggravated ..as we think it is, it is just more disseminated (broadcasted) than realized (really taking place) (P3, p.131-134). Mateus' words highlight the conceptualization of violence as a spreading disease, aggravated and disseminated as a contagious disease (like an uncontrolled force) which is thought to be more pervasive than what it really is.
The pervasive idea prevailing among the focus groups participants is that the media spreads violence by "manufacturing" and disseminating it. As a result it makes its products to be seen as contagious agents. The fact that media trivializes violence by its sensationalism and excessive show of violent scenes which are repeated over and over again on television, causes violence to acquire a common-place status and exacerbates its presence in everyday life. Violence comes, thus, to be conceptualized as an uncontrollable and contagious disease which, as an uncontrolled malignant force, is anywhere and everywhere. It also takes its toll on the way people react towards one another and on how they get affected both emotionally and psychologically.

RESULTS
Based on the MetVs which emerged in the participants talks, 05 SysMets were proposed, however, due to space limitations, only two were analyzed.
The analysis of the two SysMets proposed -VIOLENCE IS A PRODUCT MANUFACTURED BY THE MEDIA and VIOLENCE IS A SPREADING CONTAGIOUS DISEASE highlighted the role played by both discourse and cognition in the emergence of a violence representation which is embodied and social-culturally situated. One which emerges from the participants' everyday experiences and the types of information and scenes they are exposed to. In this regard, it is interesting to note that the Discourse Topics which were most salient in the two FG talks were: Media/TV (17 occurrences), Response to Violence (70 occurrences); Location (47 occurrences) and Change (22 occurrences). These at times merged into one another such as for instance when people spoke of the influence of the Media/TV on the spreading of violence and started narrating about places where violent actions occurred or when they spoke of their feelings and uncertainties, that is, how they responded in face of the prevalent state of urban violence they have to live with.
The analysis followed a bottom-up (discourse-cognition) and a top-down (cognition-discourse) procedure. Starting from an analysis of the discourse produced by the participants, discourse topics, metaphoric vehicles and metonymies were identified and analyzed. These were then grouped into families based on semantic proximity of the topics, words and expressions. Reflection based on the data allowed for the proposal of systematic metaphors as a further methodological step. Once evidence in support of the proposed metaphors was identified, a next step was to try to identify possible cognitive sub-models which are part of the representation of violence shared by focal groups participants. The analysis of the two SysMets presented here, signaled the presence of three cognitive sub-models: a FRAME schema which involves agent, patient (victim), instrument, effect and damage, and two imagetic sub-models -the CONTAINER schema which involves and interior and an exterior and the FORCE schema which involve powerful intensity and pervasiveness.
While the frame sub-model is an encompassing one and overtly or covertly is present in both metaphors, the CONTAINER and FORCE schemas are linked to the first SysMet: VIOLENCE IS A PRODUCT MANUFACTURED BY THE MEDIA, this is so, since violence is felt everywhere due to its spreading factually or to its exacerbation by the Media. This makes possible for the emergence of violence as a spreading force which in turn, has effects on people's attitudes and behaviors. So much exposure to violence makes them prone to violence which wells up in them and might be unleashed over innocent people in the streets, the traffic etc.
The FORCE schema is again present in the second SysMet VIOLENCE IS A SPREADING CONTAGIOUS DISEASE. As an uncontrolled force which carries with it something bad to be avoided, to withdraw from (a contagious disease) violence is thought of as spread out, something of which you must escape, fight off and keep your distance from.

CONCLUSION
The influence of the Media/TV in the present state of violence in urban Brazil, is strongly felt by the participants of the two FGs. The fact that the Media trivializes violence and excessively shows violent scenes on TV which are indiscriminately repeated over and over again causes violence to acquire a common-place status and exacerbates its presence in everyday life. It is A PRODUCT MANUFACTURED BY THE MEDIA which is intensely and uninvitedly offered people. Violence is also A SPREADING CONTAGIOUS DISEASE, whose force is uncontrolled and affects people in general. These metaphors, however, seem to emerge both as a result of the dynamics of talk but also because of participants embodied and social-culturally shared experiences. This is in line with the violence representation which emerges from their interaction in the stretches of data analyzed. One which presents violence with the impetus of an uncontrolled force which takes its toll on the way people react towards one another and on how they get affected both emotionally and psychologically. Revista da Anpoll nº 37, p. 293-316, Florianópolis, Jul./Dez. 2014 [43][44][45][46][47][59][60][61][62][63][64][65][66]