"- Only deal with theory after you are sixty!". I was only 23 and wanted to write about the concept of mentalities, but my Historical Method professor was too sarcastic. I asked him, playing the fool: "Should I wait because then I'll be more intelligent?". He answered "yes," but with reservation: "Or you may perhaps be losing your wit, which would be good, for the friends you will criticize will pity you and forgive you more easily." I am not yet sixty.

Sound theoretical and historiographical debate depends on the maturity resulting from extensive reading and insight to identify the most relevant authors. They may be colleagues or friends when dealing with a theme we also practice, which is the present case. In Brazil, historiographical and theoretical criticism is uncommon, so that approving reviews and indirect criticism that use the indefinite pronoun "certain" ("certain historiography," "certain authors") are frequent. Citing authors critically seems unethical. Why I think differently: having a specialist analyzing our work at length--without favoring it due to friendship or bashing it due to enmity--what else can a scholar who values debate wishes for?

To consider yourself a specialist capable of judging someone else's work sounds presumptuous. I claim I have researched the military dictatorship, Brazilian historiography, and Historical method and theory for the past thirty years.

The present article aims at identifying some of the most relevant work on the military dictatorship, whether recent or not, to discuss the following theoretical-conceptual and historiographical questions that emerged lately:

1) accusations of revisionism by Marxist historians to researchers that

1.1) label João Goulart a putschist,

1.2) question the democratic character of the left in the 1960s,

1.3) utilize the concept of political culture and

1.4) do not utilize the concept of social class;

2) the problems found in the criticism originally produced by Daniel Aarão Reis Filho to the comfortable memory on "armed struggle";

3) the importance of the notion of feedback between "armed struggle" and repression to understand the difference between the insertion of organized personnel and sympathizers in armed actions;

4) the debate on denominating and periodizing the coup and the regime; and

5) my understanding on the end of dictatorship as a long-duration process suffering little influence from opposition sectors.

Until the 1980s, theoretical debates on Latin-American military regimes, including the Brazilian regime, were concerned with general models of interpretation and took place in the field of Political Sciences. Lacking dialogue with empirical evidence, such debates never interested historians (FICO, 2004, p. 32). Available information about the dictatorship came from the press, official statements, and testimonials or memories; only recently did scholarship have access to large documentary funds.

The scarcity of sources is no longer a problem: the Brazilian military regime preserved several archives, several of these open to public consultation since the end of the 1980s - an abundance that suggested to me the term "documented dictatorship" (FICO, 2008a). Consequently, the number of empirical discoveries rose, changing the landscape of academic production in the field. If previously the challenge was obtaining sources, now it is analyzing, establishing new interpretations and explanatory hypotheses - evidently the hardest part of producing knowledge and, therefore, the one that has raised discussions.

The most controversial discussion is how researchers such as Argelina Figueiredo, Daniel Aarão Reis Filho, Jorge Ferreira, and Denise Rollemberg have been accused of revisionism by Marxist historians. In 2004, when asked if I would promote a conference on the 40-year anniversary of the 1964 coup, I suggested a reunion of all institutions in Rio de Janeiro for one big event. The idea was well-received, and the conference attracted interest from the public and the press. It was in this context that the "revisionist operation" became "dominant in the Brazilian academic mainstream," as Demian Bezerra de Melo, one of these Marxist authors would say ten years later (MELO, 2014, p. 161).

1) Revisionism

1.1) Goulart's Putschism

In 2004, Caio Navarro de Tolero, the precursor of studies on the Coup of 1964, called attention to the frailty of accusations against Goulart of plotting a coup. Opposing authors that had characterized the ex-president so, Toledo accuses them of revisionism - a grave sin that both Caio and Demian use with little caution.^[2] Toledo correctly affirms that no documents reveal any "continuist or coup plans from Goulart" -- not to mistake a coup for leftism, which accusation could be made to certain sectors from the left (that is, including Goulart and Brizola) due to their "strident words and eloquent discourses" (TOLEDO, 2004, p. 37, 48 and 44).

Debate on the supposed coup plotting character of João Goulart is interesting for there are two moments in which the then president utilized strong pressure mechanisms, spreading general fear and giving margin to speculations on eventual unconstitutional plans he might have had. He succeeded on one occasion and failed in another. I refer, naturally, to anticipating the plebiscite on parliamentarism and request for a State Under Siege. Those facts are well-known and do not require description; the facts call attention to themselves and eventual coup-plotting attempts by Goulart. It is worth highlighting, initially, that the objective, in both cases, was to lead the National Congress to vote by the government, so that it is possible to speak of undue pressures instead of coup attempts.

Historian Marco Antonio Villa (also accused of revisionism and antagonized by the left and academia due to his harsh criticism to both as a polemicist and pamphleteer on radio, TV, and newspapers) best captured such attitudes that could be regarded as foreshadowing a coup. As a historian, Villa can control a significant amount of information, using extensive biographies and press, making use of a privileged memory, with which he seasons his chronological narrative with tasty facts, giving it color and interest. Villa highlights that Goulart pressured the National Congress in 1962, aiming at anticipating the plebiscite using a note from the commander of the III Army in which General Jair Dantas Ribeira said that the uncompromising opposition of more passionate politicians could "have unforeseen consequences to the fate of institutions." Dantas Ribeiro, in a telegram to the Minister of War, said that his alert was not a threat or imposition, but a warning. The general had support from Osvino Alves, commander of the I Army, as registered by Villa, who also notes the general strike of the General Command of Workers (Comando Geral dos Trabalhadores - CGT) and the renounce of Brochado da Rocha's cabinet on the eve of the voting the anticipation proposal. Carlos Lacerda classified such pressure to anticipate the plebiscite as a coup the government was plotting, especially through the "dark cabinet," a secret cabinet formed by Jango's advisors identified as communists by Lacerda (Raul Ryff, Alvaro Vieira Pinto, Cibilis Viana, and others). Correio da Manhã condemned Dantas Ribeiro on 15 September 1962 for being unable to contain "the people's mood" should the anticipation not be approved, and called his declaration a "subversive note." Indeed, word in the press was that Jango intended to, at least, establish "absolutist presidentialism," and that his final play would come when the plebiscite was voted.^[3] In any case, political games, threats, and failed predictions were widespread, but not enough to demonstrate the existence of an atmosphere of effective coercion under which the Congress would have approved anticipating the plebiscite (VILLA, 2004, p. 82-84). That is, the Congress could have voted against the anticipation, but it is certain that Goulart went all out.

Goulart's attempt to declare the State of Siege is also weak as an indication of a coup. Indeed, as pointed by Villa, on the day he forwarded the message to the National Congress, tanks moved to the Ministry of War, but that can be justified in several ways. Marco Antonio affirms that, with the State of Siege, Goulart could interfere directly with the states and planned to do so in Sao Paulo, Rio de Janeiro, and Pernambuco (VILLA, 2004, p. 119). The problem is that the State of Siege did not explicitly allow an intervention on the states and Villa does not present evidence for intervention plans in these states, but rather he presents inconclusive or questionable proof, such as a testimonial from José Serra from 1980 and an interview he gave in 1994. In 1980, Serra - president of the National Union of Students in 1963 - assured that intervenors would be nominated in all states. In his 1994 interview, Serra claims that the State of Siege was presented as a coup in a meeting in which he participated. But in the memoir, he published in 2014, Serra relativized such claims to a great extent. He revealed that Jango confided to him in October 1963 that he knew he would not finish his term, that possible putschist repercussions were being considered in fact by Brizola, that the intervention in Pernambuco was just a suspicion he had, that Jango was not a coup plotter, and that he did not believe that Goulart "acted enthusiastically in plotting the way from the State of Siege to the State of Exception" (SERRA, 2014, p. 19). There is also a poorly told story about a lieutenant colonel supposedly appointed to arrest Carlos Lacerda. The most suspicious movement of the troops on that day took place at Palácio Das Princesas, in Recife. In his latest book, Villa reiterates the accusation of coup plotting and amplifies it, assuring that Goulart "dreamed of reelection," again without presenting empirical evidence (VILLA, 2014, p. 27).

The specialized literature does not mention an interesting aspect: if Jango was so clearly plotting a coup, how come the opposition did not seek to impeach him legally? National Democratic Union (União Democrática Nacional - UDN) deputy Aliomar Baleeiro tried so through links with the Social Democratic Party: if Jango were removed from office as of August 1963, Congress would elect his substitute, so that the biggest party would likely conquest the presidency of Brazil. The thesis did not prove to be true, and the right-wing was the one to put forward a coup.

1.2) Non-democratic Left

Concerning historians who affirm that left-wing sectors were not democratic, Caio Navarro de Toledo timely highlighted that basic reforms proposed "amplifying the excluding liberal democracy" (TOLEDO, 2004, p. 48, note 38). The question of who supported democracy was simple. In fact, Argelina Cheibub Figueiredo in her Ph.D. thesis in Political Science from the University of Chicago in 1987 affirmed that "leftist and pro-reform groups sought such reforms at the cost of democracy [...] and proposed and were willing to support non-democratic solutions. They accepted the democratic game only when compatible with radical reform" (FIGUEIREDO, 1993, p. 202). Such anachronic statements derive, perhaps, from the author's training as a Social Scientist and her MA in Political Science: her thesis lacks a historical perspective and a sophisticated theoretical foundation, as I suggested elsewhere (FICO, 2004, p. 50). Her conclusions indicate that Goulart did not know how to make adequate choices so that the opportunities for implementing reforms under democratic rule were reduced. Naturally, it is possible to notice in these conclusions the intellectual influx of Adam Przeworski, Argelina's supervisor, developer of a Rational Choice Theory. Such anachronism consists in expecting those who lived in the 1960s to be adept to democratic values, only commonly valued decades later. Jorge Ferreira agrees with Argelina's position and has also been criticized by Marxists (FERREIRA, 2003, p. 400; FERREIRA, 2009, p. 188). Denise Rollemberg avoided anachronism affirming that the left, "similarly to the rest of society [...] did not hold democracy as a supreme value", and highlighted, correctly, that "the construction of memories of such a past has been done less under the light of the values that guided the struggles at that time and more in function of the present, of the 1980s, when the reference was democracy instead of revolution" (ROLLEMBERG, 2003, p. 47-48). His last claim, however, is somewhat superfluous, for every construction of memory obeys this influx. Well, historians, if they will, may journey through the field of disputed memories, of how society appropriates the past at their own risk, but the use of the past itself, the memory that prevails, in the end, is after all beyond the scholar's control, and we can only act on it as citizens.

Critics of so-called revisionism attribute excessive importance to Argelina Figueiredo's work. Demian Melo argues that Figueiredo's thesis should be considered "a landmark in the revisionist literature on the coup" (MELO, 2006, p. 121, emphasis in the original) especially since it opposes René Armand Dreifuss (whose thesis from 1980 would be published in 1981 under the title of 1964: A Conquista do Estado), and Marxist historians hold Dreifuss' work in high regard. Incidentally, just as Argelina's thesis, Dreifuss' --whose qualities are widely recognized --is also overvalued by Marxist critics, despite its shortcomings.^[4] In the case of Argelina's thesis, besides the aforementioned absence (in my viewpoint) of a historical perspective, I believe that there is also a certain simplicity in the supposition that Leonel Brizola's radical left could simply consider rationally as a possible choice not to behave according to what they actually were, namely radical and leftist. The same could be said about Jango, that is, one cannot disregard the whole historical-political trajectory and Goulart's trajectory and simply suppose that he had an array of alternatives to choose from: several "unpleasantries" --to mention Adam Przeworski's focus-- would have to be considered.

In fact, Argelina's thesis-- besides its empirical hypothesis --more than opposing Dreifuss intends to criticize previous works that sought any structural determination to explain the 1964 coup. Therefore, besides confronting authors that assumed the inevitability of the coup due to economic factors, Argelina also criticized the determination of institutional factors, present in Wanderley Guilherme dos Santos's thesis. Santos's thesis was being prepared to be publicized since 1969, and he was possibly Argelina's most important interlocutor since they shared the same intertextual scope in North-American Political Science, with its discussions about institutional impasses, choices, etc. Wanderley's work proposed the famous thesis about decision paralysis, based on the theoretical model of "political competition and calculus of conflict." Also with an anti-Marxist vein, Santos attempts to explain in institutional terms what he calls the collapse of the political system in 1964. I do not intend to go back to this old debate, ^[5]but I would like to situate that critics of the so-called revisionists should face the real opponent - and its inextricable theoretical construction - instead of the house of cards they take as a fortress.

Curiously, it is important to highlight that both Argelina Figueiredo and Wanderley Guilherme dos Santos dialogue with famous political scientist Alfred Stepan. Argelina agrees with Stepan's statement that Goulart, in the final moment of his government, radicalized his position and lost any possibility of support. Santos agrees with Stepan's idea of fragmentation of political support but says that it did not derive from electoral coalitions, but rather from parliamentary coalitions. Alfred Stepan's book (STEPAN, 1971) is known for such general claims and, above all, its thesis that the singularity of the 1964 military intervention is in its departure from a "moderating" standard that had prevailed until then. That is, the Brazilian military would briefly intervene, "do some house cleaning" and hand it back to civilians. According to Stepan, the situation in 1964 was different: they came to stay. Why did the moderating standard change? As taught by the North-American political scientist, such a change occurred because the military felt directly threatened by the possibility of Goulart carrying out a coup, which Stepan euphemistically called "put an end to the crisis" (STEPAN, 1971, p. 143). Stepan affirms that Goulart "hoped to mobilize the political power of the masses and demand reforms through a plebiscite or a decree, pressuring the Congress or even closing it" (STEPAN, 1971, p. 141). This would have led to a change in the "moderating" standard so that the military would have opted for a long-lasting authoritarian military government. Stepan complements that, based on the National Security Doctrine, the military felt able to conduct such an endeavor (STEPAN, 1971, p. 137). Therefore, it could be said that Stepan was the first academic to consider Goulart a putschist.

Alfred Stepan's speculations became respected and referenced in academia. The book is his doctoral thesis, presented to Columbia University in 1969 with the title Patterns of Civic-military Relations: the Brazilian Political System was published as a book with the title The Military in Politics: Changing Patterns in Brazil. Dating the presentation of the thesis is important since this thematic was being discussed by the Department of State even before 1964. Indeed, as I have shown in previous works, the North-American ambassador in Brazil was able to convince the Department of State that Goulart was planning a coup. In his telegrams to Washington, Lincoln Gordon supported the thesis that the coup would be carried out with the communists' support, who would eventually seize control of everything (FICO, 2008, p. 90). Secretary of State Dean Rusk adopted on April 2nd, 1964 the official version of the North-American government, Goulart's supposed intention of remaining in power (FICO, 2008, p. 130).

Few days after the coup, the section counselor of the US embassy in Brasilia, Robert W. Dean sent a telegram to the Department of State recommending the reading of an essay he had presented to the War College entitled The Military in Politics in Brazil (coincidentally like the title of the book that Alfred Stepan would publish years later). Robert Dean argues in this essay that "the military has been developing a stabilizing role in the political life of Brazil." As the last guarantors of public order and constitutional processes, the military had a moderating influence".^[6] Robert Dean talked about this theme with Golbery do Couto e Silva in March 1965, when the general agreed with his ideas, affirming that the military's permanence in power for a longer period was important to avoid future military intervention.^[7] Still in 1965, in November, marshal Castelo Branco received ambassador Lincoln Gordon for an informal talk in which the North-American ambassador mentioned the military's role as a moderating power.

Anyone familiar with Brazilian history would know that the political role of the Armed Forces was to intervene to put the nation back on the path to order and progress just as it was about to descend into chaos, exercising the same "moderating power" present in the imperial constitution of 1889, but not to govern the country themselves. Now there seemed to have been an indication of a desire by the Armed Forces to take governmental responsibility to itself.^[8]

It is not known if Alfred Stepan was inspired by ambassador Lincoln Gordon's interpretation, discussed by Robert Dean in his monograph, about the military's "moderating power," but the similarity between both perspectives is striking.^[9] As it is, such generic evaluations rose to the status of objective postulates with a claim to truth, admitted to and referenced by academia, even if lacking solid empirical evidence and having a non-academic origin -- not to say suspicious. As a North-American, it is not far-fetched to imagine that Stepan had talked to the ambassador and the advisor. On April 10th 1964, the ambassador suggested to Secretary of State Dean Rusk that he should spread the news in the press ("additional unattributed background for the press") about Goulart's putschism (FICO, 2008, p. 140): it is likely that Gordon did so in Brazil as his boss advised. Stepan was in Brazil at the time of the coup, and was not yet known as the political scientist he would become and work as a correspondent for The Economist (STEPAN, 1971, p. 141). He published an article there on April 4th 1964 entitled Mend or End, in which he argued that Goulart intended to amend or dissolve the Constitution and the Congress, that his message to Congress had extra-legal elements, that the president could end up ruling by decree, and that the left (and Goulart himself) intended to postpone the presidential election. According to Stepan, one of the main basic reforms would be the elimination of the clause of Jango's ineligibility.^[10] That is, by coincidence or by design, he did what Lincoln Gordon wanted him to do.

1.3) Culturalism

Accusations of disregard for democracy and putschism by Goulart are empirical questions. The latter would be solved, for instance, if a plan by Jango to remain in power came up. The criticism from Marxist historians towards so-called revisionism gains theoretical depth when it points out the fragility of the concept of political culture, an aspect better developed by Marcelo Badaró Mattos. For him, the predominance of the concept of political culture - correctly identified among the authors accused of revisionism - would indicate a determinism opposite to Marxist economicism, namely, culturalism. If it is easier to verify the presence of economicism, sustaining the hypothesis of culturalism is a more complex task. Just as it is difficult to find an author in historiography that fits the economicist caricature described by anti-Marxist critics such as Glaucio Ary Dilon Soares or Wanderley Guilherme dos Santos, it is unlikely to find historians who support that the cultural sphere determines every other instance of reality without mediation.^[11] It is true that, in the moment of affirmation of Nouvelle Histoire (also with an anti-Marxist tendency), Roger Chartier also made irresolute statements in this sense.^[12] But, among the Brazilian historians here discussed, balanced approaches prevail.

It seems that the criticism that needs to be done to the concept of political culture must aim at its immanent fragility, and not at authors' risk of taking the cultural sphere as a determinant in monocausal terms. Indeed, this concept has been employed in a problematic way.^[13] Many of us use it without any problematization, referring it to Serge Berstein's basic considerations, which also stem from the criticism of the shortcomings of Marxism - a sort of theoretical tribute paid by his generation, who lived under the academic predominance of Marxism - and inserts itself in the French tradition of methodological pragmatism: it seeks answers to the problem of a politician's motivation, something to substitute opinion polls unavailable to historians. Berstein's definitions of political culture are broad and could also serve to define political mentality, political representation, political imaginary, or any other similar concept one could fashion (BERSTEIN, 2009, p. 31). Neither does he clarify if there is a global political culture and, in certain moments, the concept nears the notions of myth and utopia (BERSTEIN, 2009, p. 32-34).

In a text, he was not able to present at a conference in Brazil in 2008, but published shortly after, Berstein writes that "political action is mostly an individual phenomenon" and that political culture "results from a cultural bath in which the individual is immersed" (BERSTEIN, 2009, p. 41). Well, at the beginning of the 21st century, theoretical questions referred as structural conditionings of individual action and reflections on the philosophical crisis of the subject - widely related to the crisis of modern reason - were already part of a vast and widely debated international literature. Thus, Berstein's plain statement on the individual character of political action generates a certain discomfort. Affirming that political culture results from an individual being in an immersion bath in culture are also confusing; it is an attempt to solve by contiguity the major problem of the relationship between subject and structure. Both statements are debatable, to say the least, and weaken the author's proposition of the concept of political culture insofar as serve as its base.

In his critiques, Marcelo Badaró associates the so-called cultural determinism to post-modernism, "the general soup in which culturalism ends up finding, admittedly or not, its main nutrients" (MATTOS, 2014, p. 70). His observation is interesting and can be related to Ciro Flamarion Santana Cardoso's critiques, which establish links between the thematic of so-called Nouvelle Histoire and post-modernism (CARDOSO, 1999). However, the motivation towards cultural studies does not derive from post-modernism, but rather from the crisis of global social theories - Marxism, with the crisis of the hypothesis of economic determination - and the relative exhaustion of quantitative or serial history (which was the way such a crisis was perceived by History predominantly in France, always in a theoretically mitigated fashion). This took place in the mid-1970s: Nouvelle Histoire's manifesto is from 1974 (that is, the three-volume collection Faire de l'histoire directed by Jacques Le Goff and Pierre Nora), and thus before Jean François Lyotard's founding book (La condition postmoderne), from 1979. As can be seen, the "anthropologization" of history that includes political history is before the irruption of post-modernism.

This point seems to take us far away from the discussion on the military dictatorship, but I count on your kindness to show its relevance. Indeed, the so-called crisis of History in our field was the manifestation of a crisis of modern reason and the philosophy of the subject. It is constituted by a series of manifestations arising from the mid-1970s, namely:

(a) literary theory criticism towards History's scientific intentions proposed by Hayden White, leading to the question of narrative;

(b) the criticism of macro-structural and quantitative history made by the new historians of the Annales and their predecessors;

(c) the ensuing option for analyses of human subjectivity and everyday life, thus generating subfields such as the history of mentalities, of everyday life, microhistory, and others, influenced, one way or another, from an anthropological perspective; and

(d) the crisis of Marxism as a global social theory, especially having in mind the fragility of the hypotheses of determination of the economic basis and hierarchization and social conflict based on class structure.

Such themes are known, and I mention them here to highlight their impact, without which a reflection would be impossible: empiricism as an escape from the theoretical discomfort caused by the crisis of History and its consequences.

It is indeed a problem that naturally does not affect historiography solely on the Brazilian military dictatorship, but it does affect it. It is glaring that, since the crisis of History (it can be situated, approximately, between the mid-1970s and the second half of the 1980s), most historians' usual resistance to theoretical debate became effective discouragement: the theoretical comfort provided by Marxism no longer existed (it is worth highlighting that Marxism was predominant in academia from 1950 through 1970, including softened versions of it, such as the current practiced by the second generation of the Annales). Thus, many historians after the 1980s turned to archival and empirical evidence as a sort of refuge, avoiding wide theoretical debate and adopting, at most, ad hoc concepts. Such are the points I want to stress: because of the theoretical crisis, we experimented strong empiricism and a significant production of concepts applicable to specific historical studies. Empiricism leads to several problems -- about which, however, there is some debate -- especially about the belief of the access to what is real through vestiges, the construction of historical narrative with a claim to truth since it is based on documents, and so on.

However, from the many ad hoc concepts - that is, that are not articulated with global social theories given their decadence - we have not reflected enough. In some cases, such concepts only formalize usual behaviors, giving them a name but hardly conceptualizing them. Such procedure is reminiscent of Hempel's intentions of establishing universal hypotheses based on "general regularities," often referring to an individual or social psychology, and which is known, familiar, and understood (HEMPEL, 1942, p. 40). Those would be Nagel's implicitly assumed laws (NAGEL, 1952, p. 163). For instance, if I say, "people did everything they could to live better," probably everyone is going to agree with me. Is this a historical law? A universal hypothesis based on general regularities? And if I say that, under the terrible conditions of modern slavery, despite the brutal violence, slaves did everything they could to live a little better, even if that implied some degree of negotiation with the white masters? Another example: if, hypothetically, during the dictatorship, someone was on the verge of being arrested and had powerful friends, would they resort to them seeking for protection? The idea of contractualism associated with slavery led Marxist historians to accuse revisionists who defended it - at the beginning of the 1990s.^[14] Rodrigo Patto Sá Motta supposes that resorting to personal bonds during the military dictatorship to avoid repression was the "traditional arrangement of political culture" (MOTTA, 2014, p. 318).

Denise Rollemberg, besides extensively using the concept of political culture, also makes use of the notion of grey zone (Primo Levi's metaphor, in a specific narrative context, adopted by Pierre Laborie to analyze the French under Vichy) and would have the merit of indicating that not everything is obvious, definable - as common view affirms (ROLLEMBERG, 2008, p. 62). In this sense, we should not divide historical agents between heroes and villains, for there would always be nuanced, ambiguous positions. It cannot be denied that historians who see in society (for instance, only oppressors and oppressed) are scarcely candidates to receive any criticism. Even religion, art, and common sense have more examples of subtle interpretations capable of detecting ambivalences. Curiously, however, when Denise analyzes the OAB (the Brazilian bar association), she condemns a supposedly reproachable attitude from Seabra Fagundes when the jurist suggested that the OAB should not emit a statement about the article in the 1967 Constitution exempting actions based on institutional acts from judicial appreciation: "How could the OAB itself remain silent on this matter?" asks Denise Rollemberg, indignantly. She accuses that they would "legalize what is illegal, or even infamous" (ROLLEMBERG, 2008, p. 77). It is not the case to discuss Seabra Fagundes's position - a democratically-oriented man - but to consider that, in this case, all is said in black and white, in a tone of accusation and indignation, without gray areas or nuances, with the accused's a duly pointed contradiction, but not his ambivalence.

Rodrigo Patto Sá Motta conducted extensive archival research in the archives of information agencies operating in universities, and his book As Universidades e o regime militar (Universities and the Military Regime) brought precious revelations (MOTTA, 2014). He made elaborate use of the concept of political culture that, usually referring to the dominant groups, would encompass traces that Rodrigo seeks in Roberto DaMatta, especially the tendency to privilege personal bonds and ties, personal authority, informal arrangements, etc. Thus, using "accommodation and conciliation games," repression in the university environment would have been softened through negotiation, co-opting, personal relationships, and the "Brazilian jeitinho" of exchanging favors on the border between legality and illegality. Interestingly, Rodrigo supposes that "under the influx of this Brazilian political culture, military governments established ambiguous and conciliatory policies, where paradoxes bordered contradiction" (MOTTA, 2014, p. 16). Rodrigo offers several examples of cases Throughout his book that he identifies either as contradiction or paradox. For instance, the dictatorship persecuted left-wing professors while keeping them at their jobs, or "whereas certain leaders worked for the success of university reforms [...] the commanders of the repressive machine intensified purges" and so on (MOTTA, 2014, p. 290). Rodrigo's book inserts itself in the best research tradition that characterizes the stage after the crisis of History, one that is anchored in the survey of empirical evidence on a massive scale. But the author's effort in building a theoretical framework that seeks to overcome the limits of the usual definition of political culture as a concept should be praised.

Indeed, not only does Rodrigo Motta propose a notion of Brazilian political culture associated with the traits that characterize DaMatta's interpretation, but he also articulates the supposed paradoxes with the notion of conservative or authoritarian modernization. It is, therefore, a complex theoretical framework. For Rodrigo, such a paradoxical nature would be the key aspect of the military regime: the paradox resides in that the regime would express "conservative and modernizing impulses that would sometimes generate contradicting actions." Thus, the success of modernizing economic policies would put at stake the conservative utopia as it would promote social mobility. "On the other hand, if taken to the ultimate consequences, the conservative program would put obstacles to modernization, for the purging of every 'undesirable' or 'suspect' […] would result in a loss […] to the modernizing project" (MOTTA, 2014, p. 288-289). As one can see, the author holds rather negative expectations for the repressive actions of the regime. Indeed, Rodrigo affirms that personal relations would ease greater risks "when, due to the characteristics of the regime, stronger repressive acts were to be expected." He also highlights that "the number of episodes in which authorities opted for moderate, conciliatory acts when they had repressive methods at their disposal is startling" MOTTA, 2014, p. 292). Rodrigo assumes that "conservative utopias" aimed at the absence of social mobility. It would be sufficiently difficult to demonstrate that the military, businessmen, conservative liberals, the upper-middle class, and other sectors that probably commune in what Rodrigo calls conservative utopias did not desire social mobility, considering all the programs opposed to it adopted by military governments such as the university reform, as well as Mobral (a literacy program), Prorural (a rural assistance program), and others included in the scope studied by Rodrigo.^[15]

On the one hand, the absurd character of the second hypothesis (elimination of all suspects) demonstrates why it was not adopted: repression was highly selective, and there was nothing contradictory in said selectivity: on the one hand, the government could arbitrate between repressing or stimulating a certain sector, deciding among several guidelines forwarded by its agencies, as in any government. Rodrigo himself gives an example when he mentions the National Committee of Moral and Civics at the Ministry of Education: its highly conservative demands were not always considered by the government, "more concerned in modernizing universities" (MOTTA, 2014, p. 289). Repressing and modernizing at the same time was not contradictory to the general secretary of the National Security Council, who recommended intensifying repression against student leaders and reforming the "archaic structure" of Brazilian education.^[16]

Repression would sometimes avoid opposing certain sectors due to the negative impact it would have on public opinion or due to its relative irrelevance. For instance, the censorship of public leisure was less concerned about the theater than with television or cinema, since theater audiences were smaller. Such selectivity was not clear to Barbara Heliodora, a theater critic who accused the regime of being contradictory, since the theater was not considered important when talking about subsidies to culture but was dangerous and subversive by moral censorship (GARCIA, 2008, p. 177). To the common view, this apparent contradiction was also in cinema, since the regime both sponsored and censored films. However, Embrafilme produced 140 films between 1969 and 1981 (with direct investment, not considering co-productions). Four of these were censored for political reasons and nine for moral and political reasons (many others were censored only for issues such as nudity and foul language, as was common during the dictatorship). Therefore, the dictatorship's impact on cinema was restricted. Also, the agency that conducted censorship (Divisão de Censura de Diversões Públicas - DCDP) had certain caution with films produced by Embrafilme, precisely to avoid the death of the national film industry (MARTINS, 2009, p. 103 and 123). Besides the natural variety of orientations of the many governmental agencies, repression was selective rather than contradictory. Indeed, radical members of the military judged paradoxical that the government granted scholarships to "student agitators" (ALMEIDA, 2015, p. 93), but that was only how governmental action processed itself, weighing pros and cons, diverse opinions, as in any government.

Besides this problematic equation concerning the notion of paradox or contradiction, Rodrigo Patto aims at overcoming "the simplistic dichotomy resistance versus collaboration" proposing a triad: resistance, adhesion, and accommodation. Concerning the accommodation games, Rodrigo mentions that

The flexibility that allowed such accommodation strategies were due, in part, to the heterogeneous support base of the military regime, comprised of sectors that favored moderation in the use of violence. However, it is also important to consider the influence of Brazilian political culture, marked by a traditional tendency towards conciliation and accommodation. (MOTTA, 2014, p. 310-311)

At this point, it seems to me that the theoretical framework of As Universidades e o regime militar crumbles: do accommodation games result from political culture or the heterogeneous support base of the military regime? If from both, to which extent does it occur? And if it results from both, the effort of constructing a theoretical framework based on the notion of political culture was of little use. The explanatory potential of Brazilian political culture decreases before the statement that flexibility was due, in part, to the heterogeneity of the support base of the military regime, having the political culture only as an influential dimension. Moreover, the existence of a general political culture in Brazil is problematic insofar as it would possibly influence the accommodation games differently, if taking the universe of labor unions into account - to mention the example cited by Rodrigo - "whose leaders came from lower social strata" (MOTTA, 2014, p. 311). Similarly, the influence of the political culture would have ceased to exist in the early 1980s during the teacher strikes in 1981, 1982, and 1984, which "represented an important political change about previous years" (MOTTA, 2014, p. 347).

Also, the aid Rodrigo seeks in "influent authors in Brazilian social thought" such as Roberto DaMatta puts aside important critiques to the model proposed by the famous anthropologist. To call DaMatta "influent" is an understatement, since he became the last "explainer of Brazil" - to use the term coined by Carlos Guilherme Mota - and could become indeed part of the common sense, of media in all spheres, as much or even more than Gilberto Freyre and the myth of racial democracy, or Sergio Buarque de Holanda with the thesis of Brazilian patrimonialism. His insights are reproduced all over and thus admitted as evidence, truisms. However, there are significant controversies that Rodrigo Motta could have tackled academically. For instance, sociologist Jessé Souza supports that "people's practical behavior is explained through mysterious 'cultural heritages' unrelated to the institutions that control our lives" in DaMatta's work (SOUZA, 2015, p. 76). Jessé identifies the accommodation games, the personal relationships, and the "Brazilian jeitinho" with the notion of social capital, affirming that usually those who possess economic capital also possess social capital. For this reason, according to Souza,

when DaMatta says that Brazil is the country of "jeitinho" -- that is, of the access to powerful personal relations that compose the "social capital" -- he is, in fact, occluding the main question to the critique of any concrete modern society, which is the unveiling of mechanisms that maintain the privileged access to certain groups and classes to impersonal capitals, be them economic or cultural. (SOUZA, 2015, p. 86)

It is evident that such critiques, among others, problematize the usage of DaMatta's contributions to construct a concept of Brazilian political culture in the terms proposed by Rodrigo Patto Sá Motta. Not to mention the stereotyped character of such notions about "Brazilians." Moreover, the idea that Brazilian society is marked by negative specificity places the country in a disadvantageous condition about other countries that supposedly lived an ideal modernity, exempt from the vices that plagued Brazil.^[17] Negativity would stem from the influx of political culture, which would lead to such practices marked by resorting to personal bonds, personal authority, informal arrangements, etc. Such is a pre-modern, backward society. However, and according to Rodrigo Motta, such practices would have allowed the softening of repression, and there lies the accommodation capable of taming what could have been uncontrolled violence. Besides living in a dictatorship, we had a society (or perhaps some groups?) marked by the "jeitinho." The negative traits of accommodation or personalism are not discussed by Rodrigo. The critical character of that process would lie in the supposed contradiction between repression and modernization, for the regime would have been "at the same time destructive and constructive." Rodrigo cautiously states that there would have been an "other side" (MOTTA, 2014, p. 7) -- that is, a positive side (I would affirm) as if there could have been something good about a regime that suppressed freedom.

1.4) Conservative Modernization

Thus, the great contradiction - as expected, considering the subtitle of Rodrigo's book - would lie in the "central paradox of the military regime, the authoritarian-conservative modernization" (MOTTA, 2014, p. 293). This concept became commonplace, reproduced by several analysts of the military regime in this exact key developed by Rodrigo Motta of the supposed contradiction between a military regime (which apparently is expected to be 100% repressive and reactionary) and a development project seen as modernizing.

We all know the origins of the concept of conservative modernization, and it is not the case to bore readers with an exegesis of Barrington Moore Jr.'s book published in 1967, when modernization theory was already being criticized. However, it is convenient to remember the analysis conducted by Moore, which with considerable degree of generalization and abstraction intended to establish universal rules based on hierarchy that defined optimal or modular cases (the "modern industrial democracies" of England, France, and the United States) and cases in which, due to a supposedly weak bourgeois impulse, a sort of "reactionary coalition" between commercial, rural, and emerging industrial classes would have been responsible for modernization policies towards an industrial society (as in Japan and Germany). "But the result [in these cases] after a brief and unstable period of democracy has been fascism" (MOORE JR., 1983, p. 5). This type of coalition could also be verified elsewhere in different ways. In cases where the arrangement became unstable, "extended periods of conservative or even authoritarian governments" could have occurred, "without however succumbing to fascism." Barrington Moore Jr. proceeds to state that some of these conservative or authoritarian governments "conducted a peaceful top-down revolution that made them advance significantly to becoming modern industrialized countries." In this context, he cautiously registered in a brief footnote "based on admittedly inadequate knowledge" that it could have been the case with Latin-American countries. However, it can be noted that the analogy with Latin-American countries -- as seen, done rather horizontally -- does not occur when the North-American sociologist discusses modernization or industrialization, but when he writes about authoritarian governments that acquired Democratic traits (especially a parliament with limited powers) and sought democracy which, in the end, established itself with instability. He mentions the Weimar Republic, 1920s Japan, and Giovanni Giolitti's Italy to then conclude: "the door to fascist regimes was opened by the inability of these democracies to tackle the serious problems of their time and their reluctance or inability to introduce fundamental structural change" (MOORE JR. 1983, p. 431, 432, and note 4).

Despite Barrington Moore Jr.'s caution and the negative focus with which he establishes this brief analogy, the idea of conservative modernization started being adopted by many analysts of the Brazilian military regime. This has been done in varied ways, sometimes explicitly referencing Social Origins of Dictatorship and Democracy or, more frequently, as evidence that does not require discussion: a dictatorship (authoritarian, conservative) that promoted industrialization and modernization despite everything. Therefore, authoritarian or conservative modernization.

At this point, I would like to let readers know that my position, opposed to the idea of contradiction and paradox, is a minority. Also, that my rejection to use the concept of conservative modernization is possibly an isolated attitude, considering how widespread the acceptance of these ideas is. Therefore, it is quite possible that I am wrong.

Whichever way, the idea of conservative modernization – and its variations and spreading – was soon adopted by several analyses, remaining to this day as an explanation of the dictatorship. Soon after Moore's book was published, Fernando Henrique Cardoso discussed the dependency issue in his famous exposition at the University of Yale in April 1971, mentioning the supposed contradiction between the "politically reactionary movement of 1964" and State modernization, the expansion of the public sector, and the country's integration with international capitalism (CARDOSO, 1972, p. 71). Marcelo Ridenti, in his acclaimed thesis, defended in 1989 and published in 1991 (RIDENTI, 1993, p. 18), uses the expression indistinctly. In 1997, Ridenti used it as indicating the contradiction between "rapid growth of the productive forces" and social injustice combined with suppressions of freedom (RIDENTI, 1997, p. 20). He further elaborated this idea in 2003, in a chapter about culture and politics from 1960-1970, persisting on the commonplace view of paradox, not adopting the original sense proposed by Moore. This is possibly the origin of Rodrigo Motta's theoretical insight. Indeed, according to Ridenti

The cultural activities of the civic-military regime also implied the conservative-modernization of education, with the massification (and degradation) of basic public education, incentives to private education, and the creation of a national system of support for graduate studies and research. There, the dictatorship found some of the main foci of resistance, which it severely repressed while offering an alternative to institutional accommodation. (RIDENTI, 2003, p. 156)

In Ridenti's work, the supposed paradox and possibility for accommodation at the universities later developed by Rodrigo were already present. A few years later, Ridenti reworked this text and qualified as ironical the alleged paradox of what he then started identifying as authoritarian, but no longer conservative, modernization. Ridenti's linguistic act of characterizing the paradox as ironic is apparently minor but important: as established by studies in rhetoric, the figure of paradox found in analyses of the dictatorship is strengthened by the figure of irony.^[18] Irony frequently leads to a type of paroxysm of distancing, as it tends to the "sorry superiority" of those who claim, "they had no idea of what was going on" about the past. But the reckless combination of paradox with irony could also lead to amazement, as seen in the episode of Denise Rollemberg's indignation against Seabra Fagundes. Similarly, Ridenti talked in 2003 about accommodation as a conformed adaptation. He calls attention to what he identified as the "public decline of Brazilian left-wing intellectuals" and accused Brazilian intellectuals who once fought against the dictatorship of adapting to the new order, in what would be the triumph of the neoliberal conception of the individual (RIDENTI, 2003, p. 159-161). In 2008, he reiterated that the authoritarian modernization offered accommodation as an alternative to opposition sectors. Such accommodation, alongside repression, could "neutralize occasional revolutionary dreams, which lived with and gave space to professional development, in which everyday reality and the bureaucratization of employment prevailed" (RIDENTI, 2008, p. 36-37). Ironic and indignant, he accused the accommodated and neo-liberal intellectual of only aiming at the street movement "sat by the bulletproof window of his comfortable cabinet with an ocean view - which he never tires of looking at - awaiting news of the latest intellectual fashions from abroad, or the opportunity to see them himself in Paris, London, or New York" (RIDENTI, 2003, p. 160).

Besides Marcelo Ridenti's elaboration which would be appropriated and expanded by Rodrigo Motta as we have seen, other references to this concept are rather rarefied. For instance, Daniel Aarão Reis Filho mentions the expression after considering the misleading conjectural analyses of the 1960s left-wing about the supposed crisis that would hit capitalism and dictatorship: "under these constructions [...] it was impossible to notice that, in the contradictory tangle of the dictatorship's policies, an emerging process of conservative modernization" (REIS, 2014, p. 61). Marco Antonio Villa talks about "modernizing economic actions" (VILLA, 2014, p. 376). Marcos Napolitano states that "the path to modernization" would be marked by the "security and development" bias (NAPOLITANO, 2016, p. 7-8).

The idea of conservative modernization has the strength of common sense: when thinking about the Brazilian dictatorship, it is easy to identify repression, conservatism, and authoritarianism on the one hand, and the "miraculous" economic growth, infrastructure, and advancements in telecommunications on the other. To common sense, a dictatorship should correspond to a period of darkness, of complete stagnation. It should be regressive and entirely repressive. Thus, journalists often ask us about the Brazilian military regime having "a good side," namely the economic growth, infrastructure works, etc. This would be apparently contradictory, difficult to understand. Well, Barrington Moore Jr.'s concept of conservative modernization does not have the issue of contradiction as its main analytical vector. Seeking to oppose the still predominant currents of the limping theories of modernization, Moore attempted to formulate a sort of neo-Marxism, in fact highly imprecise and criticized -- but valuing the concept of social class as a determining factor to discuss modernization.

Associated to the notion of paradox, the concept of conservative or authoritarian modernization becomes even more fragile than Moore's original proposition for it lacks historicity. I have been saying that "we should not reduce Brazilian history from 1964 to 1985 to the history of the military dictatorship". This statement is not just a word play with which, notwithstanding, I try to call attention to the fact that not everything that took place in those years resulted from the dictatorial regime itself. Several phenomena typically associated with the dictatorship could be traced back to earlier periods. This happened with the censorship of public leisure, political propaganda, with the Federal Council of Culture, with various formulas in the political field -- several cases that cannot be detailed here.^[19] The most commonly mentioned example is the Brazilian "miracle." However, the Brazilian economy grew above its historical average in other periods after 1940. Brazil saw elevated growth right after World War II and during the period of the Goal Plan (Plano de Metas), as well as the "miracle." On the other hand, by the end of the 1970s and during the military dictatorship, GDP growth had lower averages (VIEIRA, 2008, p. 19). That is, modernization - here understood as GDP growth and its consequences - did not result from the authoritarian regime; it had been constructed for decades. I would like to stress that I do not propose counter-factual conditionals, such as "there would have been growing even without the dictatorship." What I state is that growth during the said period did not simply result from the actions of the military governments and should be read with the long-term in view.

Moreover, it is worth remembering what Marilena Chauí said long ago:

the idea of lateness implicitly presupposes as a counterpoint a "model" of society completely performed or developed, so that history is read as a modernization process, with gradual approximation and progress towards development. (CHAUÍ, 1978, p. 27)

Indeed, it is stunning that the ideological (so to speak) dimension of modernization theories is not even considered by those who insist in adopting this heavily criticized perspective. When taken to the extreme, such a view places Brazil on a level of institutional backwardness whose causes seem to lie in an overall inability of society. This view can be seen in Ronald Schneider's work, who affirms that Brazil's "incomplete modernization process" and its "political instability" - resulting from "slow development of its political organizations and institutions" - are marked by the "fragility of the political community", by an "inadequate level of institutionalization", by the "weakness and heterogeneity of political parties", and so on (SCHNEIDER, 1971, passim). Schneider did not base his work on Moore but on North-American economist Samuel Huntington who supported such ideas and even visited Brazil for academic conferences and talked with members of the Médici administration in 1972. When interviewed then by Jornal do Brasil, Huntington referred the phrase attributed to Delfim Netto that one must first let the cake grow before distributing it. Huntington claimed that it was essential to restrict social claims to allow the economy to grow.^[20] Anyhow, before overcoming our institutional needs, we will never reach the modernity of competitive and sophisticated democracies and will be fated to alternate between an "democradura" and "ditabranda", to remember the lamentable expressions used, at the time, by another North-American political scientists, Philippe Schmitter, one of which was subsequently used by unsuspecting author of a Folha de S. editorial on 17 September 2009 (SCHMITTER, 1971, p. 392).

1.5) Classist Nature

The critique by Marxists towards so-called revisionists for not using the concept of social class was nearly a lament. For instance, Marcelo Badaró remembered that Moniz Bandeira stated that the 1964 coup was a class struggle event, and then questioned: "would more recent historiography have doubts about that?" (MATTOS, 2008, p. 262). It is a rhetorical question, naturally. Several years after raising this question, Marcelo criticized Daniel Aarão Reis Filho's analytical perspective which would ultimately inspire research and publication after it was first announced at least since 2010. According to the focus supported by Reis Filho, there would have been a major consensus in favor of accepting the military regime, capable of generating support and undermine occasional resistance (REIS, 2010, p. 182, note 1 and MATTOS, 2014, p. 88). This perspective inspired the social construction of authoritarian regimes which, for the same reason, was also criticized by Marcelo Badaró Mattos (ROLLEMBERG; QUADRAT, 2010 and MATTOS, 2014, p. 89).

In 2014, Reis Filho would elaborate his focus by stating that civilians supported or subjected themselves to the regime for various reasons. Thus, "the dictatorship in Brazil was a historical construction, also due to its long duration. It is impossible to understand it without considering its political and social foundations - multiple and differentiated as they were" (REIS, 2014, p. 128).

In the extra-textual elements of the book organized by Demian Bezerra de Melo, other Marxist authors also register brief criticism to the alleged revisionists on the issue of class. In the preface, Virgínia Fontes speaks of an "attempt to erase social classes and their struggles magically." On the flap, Renato Lemos sustains that so-called revisionists "deny heuristic validity to concepts associated with class struggle, proposing other concepts, de-characterizing society's classist nature." On the fourth cover, Gilberto Calil assures that the accused sustain a conservative, theoretically fragile view, denying such conflicts, besides not using the concepts of class and class struggle (MELO, 2014). Criticism registers absence. Thus, the question arises: why should revisionists use the concept of social class? What cognitive advantages would they obtain? How should they overcome the difficulties and limitations of the concept of social class pointed out by serene critics and admitted by self-proclaimed Marxists? None of that is discussed.

Without an in-depth debate on the question of social class in the context of the 1964 coup and the subsequent regime, this controversy - institutionally restricted to Universidade Federal Fluminense - became prisoner to the framework established by Daniel Aarão Reis Filho, whose preoccupation is identifying who supported the coup and (whether by immobility or apathy) collaborated with the regime. This purpose has objectives I will further analyze but is somehow limited, classificatory, and it does not necessarily reject using the concept of social class as it is.

When Marxist historians pointed out the lack of the concept of social class, they did not seem to be defending the need for a Marxist analysis of the 1964 coup and the military regime. At the same time, they did not present the disadvantages of considering Brazilian society of the time under the light of so-called revisionist parameters. I emphasize this point because there is indeed a central analytical problem that demands equating depending on a better understanding of the theory of social classes. It still lacks resolution despite being heavily discussed at least since the beginning of the 1980s. I evidently refer to the theoretical problem of the middle classes and studies by Décio Saes. In a widely known essay published in 1982, Décio Saes discusses the 1964 coup and states that "the ruling class would find their counter-revolutionary 'mass' in the bosom of the middle class." Moreover, he sustained that such instrumentalization was carried out by the "most archaic, backward, and conservative sectors of the ruling class: the commercial bourgeoisie and the landowners" (SAES, 1981, p. 499-500) -- a statement that obviously echoes Barrington Moore Jr.'s central idea. Some years later, João Roberto Martins highlighted the difficulties in characterizing the middle classes: "support, hesitation, and even opposition (caused by the fear of proletarianization) coexist in the lower-middle class 'mass'" (MARTINS FILHO, 1987, p. 66). Therefore, several problems arise: how to insert upper and lower middle classes in the hierarchy and social conflict of subordination and dependence relationships - typical of capitalism - that is, how do they correlate, regarding class struggle, the proletariat, and the middle classes? It is evident that this old theoretical problem (and its lack of a solution) is one of the weaknesses of the global social Marxist theory. Simply it is not enough to argue in favor of using the concept of classes without even tackling it.

However, Marxist critics do not seem essentially concerned with not using the concept of social class, but rather with the occlusion of the fact that someone was defeated in 1964. Sometimes, indeed, authors under criticism are reckless and speak in general terms of support by society and not by a part of it. From an empirical point of view, that is a mistake; from a cognitive point of view, it is relatively irrelevant. What defines the nature of the events in political history is not the support network they are equipped with, but the historical subjects that engender them. The coup was supported by many, but Governor Magalhães Pinto authorized the movement of military troops and, with the government's deliquescence, National Congress decreed vacancy of the presidential office. Politicians, military, and members of Congress carried out the coup with the support, enthusiasm, or despair of several sectors. To speak simply in terms of support is inaccurate, for there are significant differences between Carlos Lacerda's wrathful action, armed and dressed in his black jacket at Guanabara Palace, or pressures from demonstrations by middle-class women with their rosaries, institutional support from the Catholic church, consent from the OAB, outcries from the press, or the simple political satisfaction from those who agreed with what they were seeing and went to the streets to celebrate, or to set the National Students Union site on fire. My reading, if translated to Marxism mutatis mutandis would be expressed according to the formula of a wide social base of a putschist movement led by the bourgeoisie. The question of support is not exempt from controversy, for it is still better to consider empirical evidence such as public opinion polls, which strongly support João Goulart and are stronger data, or as strong as, than the support given to the coup by entities such as OAB or the Family March with God for Liberty -- which, despite having many participants, was stimulated by the Catholic church's highly diversified network, having an inch of manipulation.^[21]

The military, evidently, was always concerned with the support of the regime from sectors they could rely on. The Minister of the Army recognized that the "revolution" was losing popular support during a meeting of the National Security Council, in 1968. The Minister of Aeronautics was worried about middle-class women, present in the famous marches of 1964. The result of repressing manifestations in 1968 could be harmful: "if it is difficult to repress the youth in such demonstrations, but it is even harder to oppose mothers."^[22]

For the Marxists, the 1964 coup should be characterized as an offensive action of the ruling class against the proletariat and, for this reason, "an episode of the class struggle." Indeed, the coup did not affect only labor, as sustained years ago by Maria Celina D'Araujo, when she highlighted that "the coup was against Jango-affiliated labor and against the alliances and compromises PTB made with the civilian and military left-wings." Maria Celina was followed by other authors, with variations (D'ARAUJO, 1996, p. 140; FERREIRA, 2003, p. 400; NAPOLITANO, 2015, p. 194; NAPOLITANO, 2016, p. 53 and 66). It is evident that as a movement aiming at guaranteeing and deepening Brazil into capitalism - as obvious as it sounds - the 1964 coup and the subsequent regime affected more than labor and communism. It affected all that are subject to subordination and dependence relations characteristic of this mode of production under capitalism, according to a Marxist point of view. That is, the coup affected the proletariat. In this sense - still from a Marxist point of view - the Brazilian State, before and after the 1964 coup, had a classist nature.^[23] Therefore, it is not about recognizing the classist nature of the 1964 coup and of the military regime, since the classist nature of the Brazilian State was never effectively at stake (despite those who thought a pre-1964 communist revolution possible).^[24] However, blurred by the discussion on the acceptance (support) of the coup and the dictatorship as raised by Daniel Aarão, Marcelo Badaró claims that the "classist aspect of the coup" should be recognized. What do we have, then? On the one hand, Daniel states that the coup and the dictatorship received support from important sectors of society (as Fernando Henrique Cardoso, Décio Saes, René Armand Dreifuss, and other researchers have been saying since the 1970s). On the other hand, Marcelo Badaró argues that the coup and the dictatorship are victories of the bourgeois, as no one ever denied. Known and non-contradicting realities.

Badaró admits the political dimension of such a debate: "I am moved by an eminently political preoccupation." For him, the so-called culturalism disqualifies political references of the left-wing (MATTOS, 2014, p. 93 on). It is a legitimate issue also present in Demian's texts when he states his intention to problematize "the public use of historical knowledge understood as parts of the dispute for hegemony" (MELO, 2014, p. 157). However legitimate, the issue strikes me as superfluous. To quote a famous Marxist, Eric Hobsbawm once wrote that "professional historians produce the raw material for nonprofessionals to use or misuse" (HOBSBAWM, 1998, p. 285). We certainly do not have control over social uses over the knowledge we produce. However, we have a responsibility about the theoretical density of the debates we propose.

2) Shifting sense

The old question of the tension between memory and history appears in Marxist criticism mentioned as an alleged social responsibility that historians criticize would have in fragilizing leftist perspectives insofar as they fuel rightist discourses. Among historians accused of revisionism, this issue concerns the supposed necessity that historians have of pointing out the truth covered by the prevailing memory of the dictatorship - an equally difficult task.

Indeed, there is a basic problem in the known perspective proposed by Daniel Aarão Reis Filho that can be easily distinguished in work inspired by him, The Social Construction of Authoritarian Regimes. It is a critique on the prevailing memory and the supposition that such a predominance had the support of academic production. At least, that is manifested when looking at the two initial motivations of the organizers, Denise Rollemberg and Samantha Viz Quadrat, registered in a presentation entitled "Memory, History, and Authoritarianism" - published in all volumes.^[25] The first motivation is an estrangement: Denise and Samantha register how frequent is "the construction of memory that claims that authoritarianism was only possible due to coercive and manipulative practices and institutions." The second is the authors' assumption that "[...] little has been researched and written in the sense of understanding the relationships between society and the regime inaugurated on March 31st" (ROLLEMBERG; QUADRAT, 2010, p. 11-12). There is nothing strange about the emergence of comfortable memories after traumatic events. It is a natural reaction of society, tending to be one of the most common. Nevertheless, traumatic memory and frustration also occur. As it is, it is the task of historians to unravel the genesis of such a complex phenomenon. A perspective that states that the construction of such memories obeyed unmentionable designs is unfruitful, even because such processes did not obey intentionality parameters. But, if the initial estrangement - however, justified - concerns memory, the second statement refers to academic research, which would not have paid sufficient attention to the relationship between society and dictatorship. I will not bore readers with a long list of pieces of research dealing with this issue, since Denise and Samantha are possibly expecting more research on those who supported the coup and dictatorship, in the case of Brazil. My point is not about this lapse, but about the fact that we do not have control over the social use of our production, as I have already mentioned. In fact, it is highly pretentious to suppose that academic production generates any comfortable memory, it is more likely that it follows other influxes. In the case of Brazil, when according to Daniel Aarão Reis Filho a "shifting of sense" took place, generating such a memory. Academic production on dictatorship was incipient. In fact, Daniel himself suggested that the "notable reconstruction" took place in an amnesty context, even to incorporate political prisoners accused of blood crimes, who, therefore "deserved amnesty" and could "claim moral and material reparations from the State." Daniel does not explain the persistence of this shifting of sense throughout the various conjectures that stretch from the Amnesty Law in 1979 until the moment the Amnesty Commission was created in 2001 to provide material reparation (REIS, 2004, p. 49).

Denise Rollemberg explored this issue more thoroughly. In the book The Social Construction of Authoritarian Regimes, she confronted the memory built about the Brazilian Press Association (Associação Brasileira de Imprensa - ABI) with research conducted by this institution which became highly critical about the dictatorship approximately after AI-5, despite certain omission at of the coup. Denise's narrative is enriched by irony, hyperbole, and emphases -- resources that grant it liveliness and radiance but lead to inaccuracy or generalization at times. Thus, when describing the comfortable memory, she accuses "society" of building "the image of itself as essentially democratic, rejecting tyranny from the beginning, in an intransigent struggle." Similarly, she affirms that this collective memory "forgives" those who supported the regime (ROLLEMBERG, 2010, p. 2 and 131, emphasis in the original). However, the problem is not in the correct identification of the phenomenon of comfortable memory, nor in historical research that identifies agents and institutions that supported the regime, but in focus based on procedures of accusation and guilt.

Daniel Aarão's idea of shifting sense is interesting precisely because it deals with this comfortable memory. The use of the word shifting could be discussed, for there lay assumptions on what kind of sense existed before and shifted -- when it is the construction of original memory de facto. As it is, Daniel's critique is known: he understands that the left wings who thought themselves revolutionary came to see themselves (and be seen) as members of a democratic resistance in the amnesty context -- when they had originally been supporters of a revolution, not of democracy.^[26] Daniel Aarão observes that this understanding ended up fixating itself as "irrefutable truths, objective historical processes, and not versions considered appropriate by their authors" (REIS, 2014, p. 133). He identifies the sources of the shifting. One source is Fernando Gabeira's book, overrated and derided by Daniel as a "sympathetic and good-natured version" of the "guerilla saga" that would have been capable of generating an "atmosphere of indulgence and understanding" (REIS, 2014, p. 9). Gabeira's book was published in 1979. Eighteen years later, filmmaker Bruno Barreto freely adapted the book into a film, without being faithful to the facts and the people involved in the kidnapping planned by Franklin Martins and Daniel Aarão Reis Filho, directors of the Guanabara Communist Dissidence (SILVA, 2009, p. 214). The film ended up being nominated for Best Foreign Film at the Academy Awards but was not received positively by former leftist militants and other critics. Daniel organized the publishing of a book collecting articles to criticize the film (REIS FILHO, 1997). For him, Gabeira's book was successful because "most readers" desired the "distanced, critical, ironical attitude" of his narrative. Referencing Gabeira's (and Zuenir Ventura's) book as ironical, critical, and distanced, Daniel affirmed that "these authors were the most refined expression of their time [...] is it that important that they made mistakes in narrating their stories?" (REIS FILHO, 1997, p. 36). Apparently, it does matter, considering the amount of criticism they received. As it turns out, society was predisposed to welcome comfortable versions, in Daniel's view: "Other memories [...] did not deserve the same welcome for they were not in tune with general aspirations of appeasement and conciliation" (REIS, 2014, p. 9). Gabeira's version would be the one about conciliation (REIS, 2008, p. 16).

The publishing of Brazil: Never Again is also seen as a source of comfortable memory as it exposed the aggressions, tortures, and supposedly elaborated a victimization discourse: "Brazilian society lived the dictatorship as a nightmare that needs to be exorcised" (REIS, 2014, p. 13).^[27] This memory would become, after all, prevailing and candidate to the official history of the dictatorship by means of the project Revealed Memories.^[28] According to Daniel, "the 2012 constitution of the so-called National Truth Committee is also part of historical reconstruction" (REIS, 2014, p. 13). However, at certain moments Daniel chooses to accuse Brazilian society of not confronting the past, omitting itself: "strictly speaking, the Brazilian society of the time [of the amnesty in 1979] chose silence as memory, as done often. In other words, silence as a form of memory" (REIS, 2014, p. 136).

The statement on a predisposed Brazilian society towards this discourse or memory remains to be proven. Similarly, the positive reception of Fernando Gabeira's book should not have been motivated solely due to the supposedly (and debatable) conciliatory version it may contain. Moreover, the English Marxist historian observes that many of the ideological debates we have "concern matters about which non-historians know little and care less" (HOBSBAWM, 1998, p. 285). Anyhow, the narrative of Four Days in September evidently has qualities, beyond the curiosity the theme arouses. Contrary to what Daniel affirms, other memories were also heartily welcomed, such as the best-seller The Carbonari (by Alfredo Sirkis) whose influence can be measured by the fact the author granted Rede Globo the right to use some passages to be used in the soap opera Rebel Years (SIRKIS, 1998, 31). The prevalence of the comfortable memory, according to Denise Rollemberg - who agrees with Daniel Aarão about Gabeira - would have been capable of condemning "dozens of memories" to oblivion, including the military "who lost the battle over image" (ROLLEMBERG, 2003, p. 76). This requires clarification as to which memories these are and how the military is always positively evaluated in opinion polls. Finally, there is the issue of silence -- common in situations such as these, known phases such as suspension or latency about which I have dealt in another work (FICO, 2013).

Some of Daniel Aarão's main premises are inaccurate and lack empirical evidence. For instance, he observes that "most of the Brazilian society preferred to demonize the dictatorship." Or, that "to elaborate national conciliation and a democratic pact, the prevailing version was that Brazilian society only endured the dictatorship," and that "many felt accommodated in polarizing ways of thinking" (REIS, 2014, p. 7). The lack of proof and the inaccuracy of expressions such as "the majority" and "tended to prevail" -- just as the generalization present in the expression "society" - make the statements methodologically unsafe, despite the apodictic tone. We do not know what parts of society demonized the dictatorship (probably having supported it before).

However, my reservations do not touch the main issue: the history of the military dictatorship has been seen as the history of the confrontation between political repression and "armed struggle." Such a confrontation became the emblematic fact of the period, constituting memory, imaginary, iconography, filmography, and so on, attributing urban armed actions, the so-called Araguaia guerrilla, and repression excessive historical importance. There are no events about the military dictatorship in which photos of the 1968 student manifestations are present, frequently mistaken for armed struggle. Here I digress thanks to the investment made by Correio da Manhã to its photojournalism since 1964; we have excellent visual records of the student manifestations in 1966 and 1968. The work done by Correio da Manhã - whose photographers approached the students closely - stimulated Jornal do Brasil. I would highlight the photographers Rubens Seixas and Alberto França, from Correio da Manhã, and Evandro Teixeira e Alberto Jaco, from Jornal do Brasil (OLIVEIRA, 1998). As these are the only images we have of repression, they are, often, diffusely perceived as images of armed struggle.

Anyhow, such a transformation of the confrontation of "armed struggle" and repression in emblematic fact must have been motivated by the violence and the traumatic dimension of the facts in the analysis. But, as previously mentioned - and I apologize for the repetition - Brazilian history of that period cannot be reduced to the history of the military dictatorship. When I put the expression "armed struggle" between quotes, I am not disqualifying the left that opted for violent actions but making reservations that such a process was inexpressive in Brazil. Even if it had dramatic consequences to those involved and to the very process of exiting the dictatorship - as we shall see - armed actions were scarce, and their truly offensive or revolutionary phase was rather brief since, before the obvious superiority and truculence of repression, they became merely defensive. What is relevant to the discussion raised by Daniel is not necessarily the comfortable memory he identifies, but the fact that historiography persists in characterizing the armed action as an emblematic fact of the history of dictatorship. Besides the problems, the issue insistently raised by Daniel Aarão Reis Filho strengthens this perspective and leaves aside the "conservative majorities [...] who were immobile" as the anodyne subject of history (REIS; ROLLAND, 2008, p. 13). "The conservative majorities, resigned, self-absorbed, [who] were barely at their windows, and did not even see the year [of 1968] go by" (REIS, 2008, p. 17).

3) Taboo

After the widely known 1968 demonstrations led by high school and university students, the year ended - contrarily to what Zuenir Ventura supposed - and ended badly on December 13th, 1968 with the decree of AI-5.[29] After AI-5, students frustrated by what happened could be easily recruited by self-proclaimed revolutionary organizations. These students left their families and friends, lived isolated in temporary accommodations, adopted aliases, were subject to demanding routines imposed by the organizations to keep them under "maximum tension" ^[30]and eventually participated in armed struggled and risky activities that demanded preparation without adequate training. Many were arrested and tortured; some were killed. These young men and women's affiliation with the organizations surely occurred to more immediate motivations in comparison to what made more experienced and higher-ranking militants to choose, before AI-5, "armed struggle." This is one of the persisting taboos that remain untouched by military dictatorship historiography. It is a delicate subject since it calls attention to the fact that so-called revolutionary organization leaders could foresee the tragic results of recruiting inexperienced young militants.

In his acclaimed work on the student movement, published in 1987, João Roberto Martins Filho remembers the attraction Jules Régis Debray's booklet Révolution dans la révolution? Lutte armée et lutte politique en Amérique latine - a pamphlet that simplified and mystified armed struggle - caused on those young militants (MARTINS FILHO, 1987, p. 197). But the organizations themselves also transmitted an impression of power, showing themselves capable of raising money and go on daring enterprises, besides using military jargon that seemingly made them appear capable of rivaling actual military. Even after years of self-criticism, Franklin Martins still talked about "military action."^[31] Dissatisfaction with older politicians also may have played a part in attracting students to the cause, since these politicians, according to Franklin Martins,

Had been defeated in 64 without putting up a fight. They promised the people a new life but left the people alone when it came to it. They took refuge in embassies, went underground, and chose to wait for calmer times. At least, they had been naive and unprepared. Why should the youth take their advice into account? (MARTINS, 2002, p. 45).

Vladimir Palmeira, whose performance as a student leader was impressive, when speaking 20 years later about 1968, remembered the recklessness of the students' choice. He said that students began "to prepare a plan for armed struggle, a revolutionary plan, a plan for anything -- it was all an abstraction. AI-5 would give this great strength". Asked about magnetism, so-called revolutionary organization's ability to attract members, he replied:

You had a politicized fraction [that is, more politicized student groups participating in the 1968 manifestations], participating in confrontations. Suddenly [with AI-5], there was no longer a possibility for institutional struggle [with manifestations, marches]. At the same time, you had a call to arms [from the supposedly revolutionary organizations], and the performance of victorious actions as the police took some time to adapt and start striking these organizations down. But at this point, hundreds of students had already joined the ranks of armed actions. (PALMEIRA, 2008, p. 124)

In his thesis, Marcelo Ridenti accounted that students were possibly more than half of left-wing militants in general, "particularly in those [groups] that took up arms" (RIDENTI, 1993, p. 115-117) based on data from the project "Brazil: Never more." Daniel Aarão Reis Filho situated the student movement as a "pawn in a more complex game - revolution. Since, however, it was the most active and articulate social movement, it was the one that recruited militants and leaders to the revolutionary parties" (REIS FILHO, 1998, p. 32).

In a memoir published in 1982, Herbert Daniel -- former Revolutionary Popular Front (Vanguarda Popular Revolucionária - VPR) militant, wrote that he took up arms without knowing what to do with them. A more experienced member told him that, if required to shoot, he only needed to point and pull the trigger: "That was the most complete and only military instruction I had during my early times in the guerilla" (DANIEL, 1982, p. 17). In this thesis, Marcelo Ridenti commented the repression suffered by organizations and the difficulties they faced to survive, even regarding their ranks, which obliged them to "resort to the last human resources available to fight dictatorship: their supporters, particularly students, politicized in the social movements of 1967 and 1968". In this context, Marcelo cited Vinícius Caldeira Brant's testimony (former Student Union president and former Popular Action and Revolutionary Workers' Party militant, arrested at 29):

Suddenly, I found myself in a DOI-CODI cell in 1970. In the other cells, there were a bunch of boys. [the left wing] was being decimated and the people who were only supporters until then started joining the struggle and taking risks. Sometimes, those people were not prepared. (qt. in RIDENTI, 1993, p. 253)

When I mention the inexpressive character of "armed struggle" in Brazil - regardless of its becoming the emblematic fact in the history of dictatorship - I reserved the obvious impact such an experience had on the lives of militants in general. Particularly, the arrest, torture, and death of these boys and girls made the final phase of the process a traumatic event. I do not refer solely to the individual impact this may have caused on surviving militants - who do not usually elaborate on the matter in their testimonies - but to the virtual interdiction of the subject. The theme is rarely perceived by the relevant literature, except for brief mentions to recruiting, that does not deal with its traumatic dimension. This silence is somewhat relevant and should deserve attention from researchers. The insertion of this thematic in the history of traumatic events of the 20th century, internationally and widely discussed expertise^[32], might be seen with disdain in Brazil as here a strong analytical tradition prevails. Marked by Gilberto Freyre's optimistic thesis about the singularity of the Brazilian character and by Roberto da Matta's less popular pessimistic interpretation, which claims such a singularity is marked by a "pre-modern personalism" as previously seen (SOUZA, 2015, p. 70).^[33] I believe if Freyre's and DaMatta's theses are highly questionable in their original versions, society's rarefied reception of them - nearing cynicism - leads to consecrated notions on singularity, cleverness, the jeitinho, and so on.^[34] But as previously said, both the theses and their rarefied reception are widely accepted. Maybe this favorable predisposition can explain the predominance in the specialized literature of paradox and irony, as previously explained. We would be absurd and ridiculous: saying it entertains the readership. Thus, the difficulty of seriously considering the question of trauma. Hence the supposed difficulty of understanding the country, except for professionals.

The strength of such interpretations should not intimidate the few who see them critically as a serious loss to historiography - I will put it this way to not sound grandiloquent. Despite the predominance of readings marked by contradiction and irony, we must unravel the question of trauma as a historiographical problem. It is possibly the cause of constructing the confrontation between "armed struggle" and repression as the emblematic face of the dictatorship, despite the scarce number of armed actions, their geographical and chronological restriction, taking place basically only in the cities of Rio de Janeiro and Sao Paulo (mostly in 1969 and barely happening in 1971). The hypothesis that the construction of this emblematic fact resulted from trauma caused by the violence naturally demands sophisticated research. It is surely not a traumatic memory in the same way as in the Argentinian experience, but it is a memory restricted to leftist groups and marked by frustration, as I suggested elsewhere. Frustrations with the failure of armed struggle; with the immutability of the military project of transition; with the absence of a clear rupture between dictatorship and democracy; and with the lack of punishment for torturers (FICO, 2013).

This hypothesis also seems plausible for another reason. Former military testimonies on the persistence of activities of organizations who chose armed struggle, even after its inviability became apparent, indicate that there was an awareness about such obstinateness. The uselessness of guerrilla-armed action after the successful repression soon became apparent but, at the same time, the inertial decision to give continuity to such actions marked a self-destructive predisposition. This possibly happened by the end of 1969 and beginning of 1970, marked clearly by the death of Carlos Marighella, leader of the National Liberating Action (Ação Libertadora Nacional - ALN) in 1969. The kidnapping of North-American ambassador Charles Elbrick (August 1969) and the subsequent strong repression almost catatonically worked as offensive stimulus and defensive feat. The author who better approached the self-destruction phenomenon -- but without considering the traumatic dimension, I stress -- was Marcelo Ridenti, who dedicates the last pages of this thesis to such a sensitive theme (RIDENTI, 1993, p. 266 on).

The inertia and suicidal attitude of militants, not to say banal, must have also left a mark on the memories that survivors constructed about the period. This hypothesis proves it would also require complex research. Indeed, if these hypotheses - about the impact on survivors of the violence against inexperienced young militants and the uselessness continuation of "armed struggle" in the phase that former militants themselves refer as survival - are correct, maybe we will be able to better understand the genesis of the emblematic fact and the comfortable memory of the "armed struggle". They did not result from discreet facts as Fernando Gabeira's book or pecuniary interests involved in the amnesty process, but rather from a traumatic background. Leaders and more experienced organized ranks suffered were arrested, tortured, and died. As if it were not enough, so were young and inexperienced sympathizers.

Besides the traumatic dimension marked by serious ethic-moral conflict, the distinction between leaderships, organized ranks, and sympathizers is important for another reason. The easiness of recruiting high-school and university students - frustrated with the AI-5 after the expectations created by the 1968 demonstrations and astounded by the apparent strength displayed by revolutionary organizations - situates a historiographical question I mentioned in previous works. There was no causal relation between repression and "armed struggle," despite what both former military and former militants say, as well as some analysts (FICO, 2001, p. 63-64; FICO, 2008, p. 199).^[35] The left wing did not opt for armed actions because of the AI-5. However, this concerns those who joined the organizations for, let us say, ideological reasons before the AI-5. Sympathizers were taken to organizations by the hopeless conjuncture resulting from AI-5, besides other already mentioned reasons.

In epistemological terms, it is not fair nor adequate to use "counter-factual" hypotheses. If it is already difficult to explain why the so-called revolutionary left opted for armed struggle and persisted with urban guerrilla, even before evidence of its weakness to repression, it is impossible to speculate about what might have happened if organizations had given up even before being annihilated. Still, this is an important historical question since the importance of the feedback mechanism has not been sufficiently highlighted. "Armed struggle" and repression did not have a causal relationship but had feedback in the sense that the left justified its option for armed struggle due to the increased repression, and the hardline justified repression due to armed actions.

However, the theme should not be treated under a reductive light -- that is, identifying the supposed guilt for increased repression or a puerile debate about who started it. "Counter-factual" speculation is important here because, if the organizations had interrupted the armed action, we know all too well that repression agents would seek other victims. That is how they acted when they defeated the "armed struggle" at the beginning of the 1970s (they went after non-violent communists). That is why we must not establish a causality relation. Security and information communities were created and maintained themselves independently from the existence of revolutionary organizations. However, if the organizations interrupted armed struggle, the feedback process would stop by the simple absence of one of the poles, prematurely weakening information and security bodies, so that claiming that armed struggle slowed up the end of the dictatorship.

Indeed, what is referred to as radicalization of late 1960s demonstrations corresponds to the radical left-wing strategy of utilizing students' dissatisfaction with education and food conditions (among other problems) as a way of bringing them together to the revolutionary struggle. According to João Roberto Martins Filho, "the direction of the Student movement in 1967 had reached the proposition of increasing popular struggle, aiming at overthrowing the dictatorship and installing a popular government". Student actions in 1968 had traces of "semi-guerrilla," but there had been radicalized student protests since 1966 (MARTINS FILHO, 1987, p. 176 and 185).

All these protests were considered in the implementation of the political repression apparatus. There was a somewhat objective evaluation of the losses brought by the protests, not only in political terms. Fifteen days after the March of the One Hundred Thousand, Costa e Silva met with the National Security Council to hear from its members if he should decree a new institutional act. During the meeting, the finance minister Antônio Delfim Netto said the demonstrations had caused losses of about NCr$ 43 million, since they had foreseen collecting NCr$ 168 million but only collected NCr$ 125 million in June, when the biggest demonstrations took place. The minister assured that the loss equaled the total of the irrigation program or 1/3 of the telecommunication program.^[36]

From a political point of view, the head of the National Intelligence Service (Serviço Nacional de Informações - SNI) Emílio Garrastazu Médici, who would later become president, informed the board that the demonstrations had procedures consisting of, firstly, disseminating propaganda presenting the government as dictatorial and submissive to North-American imperialistic interests and, by using the students' "fair claims", utilized the "energized under acceptable premises" mass to demoralize the police, "and to promote general subversion and a civil war". For Médici, the leftist offensive had reached an advanced state:

At the moment, what is seen on the streets is no longer legitimate student claims, but rather a movement with a clear goal -- that is, to overthrow the revolutionary government [...] to further implement the Brazilian socialist revolution [...]. There is no doubt that the movements of false students [...] represent the counter-revolution.^[37]

Médici mentioned the bank, weapons, and explosives robberies and "the utilization of students as maneuver mass" as parts of the left-wing plans to take over power. He was favorable to the immediate decreeing of a new institutional act but followed Costa e Silva's advice of only using the police. As known, in the following years, during his presidential term, Médici ordered restless combat against armed struggle. In a rare interview in 1982, he said that such confrontation was "a war we embraced" and boasted: "I ended terrorism in this country. If we had not embraced the war, if we had not acted drastically, we would still have terrorism" (SCARTEZINI, 1985, p. 36). Also for Médici - according to the memory he constructed - the confrontation between repression and "armed struggle" was the emblematic face of the dictatorship.

The July 11th, 1968 meeting of the National Security Council was as important as - or perhaps even more - the one on December 13th, 1968, which put Costa e Silva's decision forward to decree AI-5. I say so because the feedback process I referred to became clear in July. More than a simple pretext to harden the regime, the radicalization of the left wings resulted in genuine concern for some sectors. Sao Paulo people in business worried about bank robberies (and, for that reason, financed the Bandeirante Operation) and the middle classes -- including those who initially empathized with the students claiming for more social investment and those who were victims of repression. This evaluation was not a ruse, nor was it a simple pretext to harden the regime, even because the meeting did not result immediately in any institutional acts. In this sense, I call attention to the genuine concern, in as much as it expressed actual fears: the fear of the military, the businessmen, and other sectors that the manifestations would unfold from the political-institutional field into rioting or a revolution, as thought by Médici. Precisely because the AI-5 was not decreed in mid-1968, and because the revolutionary left wing and the student movement attached to it did not cease activities that could be classified as radical, is that the issue of feedback can be discussed. Such interaction took place exactly in this period, in which the main agents already knew what was at stake.

Sectors in favor of an immediate opening of the punishment season, represented in the National Security Council by the Minister of Justice among others (who, thus, avoided that the military ministers opposed Costa e Silva's known guidelines of not editing institutional acts) would not rest until they obtained what they wanted, ever since AI-2 had its validity revoked in March 15th, 1967. Face the waning of student protests; the so-called hard line promoted a series of provocation acts and attacks (that did not occur before) to perpetuate a climate of unrest, nearly forcing Costa e Silva into decreeing the new institutional act after the July 11th, 1968 meeting. A simple consultation to the chronology is enough to verify that, until the March of the One Hundred Thousand on June 26th, 1968 only social manifestations under student leadership had taken place. After the Council's meeting on July 11th, the following events took place: depredation of the theater where Roda Viva was being performed (Jul 17th), bomb attack against the Brazilian Press Association (Aug 29th), invasion of Universidade Federal de Minas Gerais (Sep 29th), invasion of Universidade de Sao Paulo (Oct 2nd), the kidnapping and beating of actress Norma Benguell (Oct 8th), not to mention the aborted attempt of using Air Force paratroopers to eliminate opposition members and perform bomb attacks (as denounced in October). All these episodes were provoked by the hardline. Such provocations and the escalating pressures by the military ministers against Costa e Silva springing from Márcio Moreira Alves's speeches have led to AI-5, which the marshal was obliged to sign.

Such provocations and attacks were the resources used by the most brutal sectors of the regime to pressure the military presidents as of 1968. It is a standard not yet pointed out by literature. Costa e Silva, for instance, knew these episodes were promoted by the radical civil and military right. The same happened to Geisel (the murders of Herzog and Fiel Filho, bomb attacks at Brazilian Press Association and the Brazilian Bar of Attorneys), and Figueiredo (several in 1980, and to Riocentro in 1981). Costa e Silva, however, did not resist the terrorist right's show of power.

4) Names and dates

Daniel Aarão Reis Filho insists in calling the coup and dictatorship "civil-military," supporting that the expression "military dictatorship" was used by the regime's opponents only to weaken its main representatives (REIS FILHO, 2015, p. 237). Opposition to the regime would have done everything to characterize the dictatorship as the military. However, when consulting documents of revolutionary left-wing organizations, we can see that they did call the regime a dictatorship, often times military dictatorship, frequently just dictatorship, also occurring "Castelo Branco's dictatorship", characterization given by the Communist Party of Brazil (PC do B); "dictatorship de facto" and "dictatorship of class", by the Revolutionary Marxist Organization of Workers; "counter-revolutionary dictatorship" and "neo-colonial military dictatorship" by the Red Wing - Communist Party of Brazil; "strong regime" by the National Liberation Command; "bureaucratic-military apparatus", "reactionary dictatorship", and "military dictatorship, expression of the landowning-bourgeois power" by the Brazilian Revolutionary Communist Party; "dictatorship of the ruling classes" by the Revolutionary Armed Vanguard of Palmares; simply dictatorship or regime and "1964 coup d'etat" (not "military coup") by the Revolutionary Movement 8th of October, among others (REIS; SÁ, 2006). In their May 1965 resolution, the Brazilian Communist Party used the term "reactionary and servile dictatorship." The variety of uses weakens Daniel's argument, giving the impression that it is an ex-post facto elaboration.

The civil-military denomination would be reasonable, according to Daniel, due to the civilian support the coup and the dictatorship had, which would constitute the "civic dimension of the dictatorial regime, even if the military chiefs occupied the top of the power pyramid" (REIS, 2014, p. 62, emphasis in the original). However, virtually, everything in politics can be analyzed according to the support and adversity network it is inserted so that saying that something had outside support does not remove its essence, its immanent configuration. For instance, Estado Novo could be considered a civic dictatorship tutored by the military. The 1937 coup could be considered military but with civic outfitting, as said by Aspásia Camargo. Getúlio Vargas would have been able to use and contain the military, according to José Murilo de Carvalho.^[38] The military's support - in this case decisive - does not remove the civic nature of that regime.

On the other hand, the civil qualifier says little when used only in opposition to the military. In this sense - as previously appointed by Demian Melo - the use of civil made by Dreifuss (with a classist, business connotation) was more precise (MELO, 2014, p. 168). Moreover, if Daniel's argument were reasonable, we would have to classify the dictatorship was civic-military-business-mediatic-catholic and so on, veering the debate even more towards nominalism.

A relative variety of uses can be found. Eliézer Rizzo de Oliveira uses the term "political-military movement" in this 1976 book; Marcelo Ridenti, in 1993, already used the expression "civil-military coup", before Daniel's more insistent positioning, and adopted "civil-military regime" in a 2003 publication; João Roberto Martins Filho preferred "political-military coup" in his 1987 classic book; Marxist historian Virgínia Fontes reinforces Dreifuss's option with the expression "corporate-military dictatorship, " and some authors prefer to leave it as it is, calling it simply military dictatorship.^[39]

The coup was effectively carried out (not just supported) by both civilians and the military. Therefore, it is possible to call it civil-military. Some subsequent steps marked the military character of the new regime, mainly the decision of the Supreme Command of the Revolution to abort, as first institutional act, the ongoing discussion in the National Congress about conceding or not to the winners’ power of expelling congressmen and, then, extending Castelo Branco's term until March 15th, 1967 (decision taken in July 1964). The military entirely controlled the subsequent regime, so that adjectivizing with reservations ("it was military, but also civic" or corporate or whatever) is superfluous and imprecise - besides having, like everything else concerning History of the Present, immediate political implications. In this case, precisely due to its adversity, the connotation reduces the military's responsibility.

This issue acquires density and interest when associated with the periodization proposals supported by Daniel Aarão Reis Filho and Marco Antonio Villa. Villa considers that we can only talk about military dictatorship during the validity of AI-5, still with reservations.^[40] In this period, "the Executive held full powers and exerted them dictatorially" - an irresolute and tautological definition (VILLA, 2014, p. 370). He supports that we cannot talk regarding dictatorship from 1964-1968 "in the etymological sense of the concept," but does not study the origins and evolution of the word dictatorship. To Villa, the period from the coup leading to AI-5 was not a dictatorship because the National Congress and other legislative houses were open, government and mayor elections were held (except in capitals and cities considered as national security areas), there was freedom of press, as well as intense cultural activity (cinema, music, theater), as well as the student movement.

It would be possible to show that every aspect highlighted by Villa is debatable since, between 1964 and 1968, severe restrictions to the functioning of National Congress (including its closing) and other legislative houses took place, besides limitations concerning elections, censorship of the press, theater, cinema, and books, as well as brutal repression against the student movement. But this would imply boring the readership with lists of examples and would miss the central point. If we remove what is tautological in Villa's definition, what is left is that there had been a dictatorship when "the Executive had full powers." Well, it can be observed that, during the period under consideration (March 31st, 1964 to December 13th, 1968), the Executive had full powers during the validity of Institutional Acts 1 and 2. As is known, the first act later known as AI-1 allowed (Firstly the Supreme Command of the Revolution and later the President Castelo Branco) to revoke mandates and political rights until June 15th, 1964. Thus, from the decree of the first act (April 9th, 1964) until today there had been a dictatorship - it could have been thought if according to Villa's reasoning. The deadline of 15 June was the limitation the victorious putschists imposed only about these two punishments (revoking mandates and political rights), but the act established another punishment: the possibility of firing, compulsorily retiring, transferring or reforming civil and military servants. This would be valid until October 1964. It would be the case of deciding if this counts as "full powers" and extend the dictatorship or not until October. The act as a whole (note that the military cared deeply about deadlines) would last until the inauguration of the president who would have succeeded the term initially held by Jânio Quadros, continued by João Goulart, swept by the coup momentarily to the president of Chamber of Representatives Ranieri Mazzili, and would finally be concluded by marshal Castelo Branco. That is, as of the decree of the Institutional Act, there was not the understanding that the military would remain in control of the situation for years and years. The act foresaw normal elections on October 3rd, 1965 and inauguration of the new president on January 31st, 1966, when the act would lose its validity as a whole. Thus, the main punishments lasted until June 1964. The possibility of firing civil servants and sending the military to reserve until October 1964 and the other prerogatives established by the act lasted until January 1966. Such prerogatives would last until 1966, giving greater power to the president and the Executive, making it easier to declare a state of siege, allowing the passing of laws by the deadline and approving constitutional amends by an absolute majority - but maybe all this cannot still be considered "full powers."

It is of utmost necessity to pay attention to such details because none of this, evidently, was carried out thoughtlessly, since it was nothing more, nothing less than the institutional framework of the "victorious revolution." Why different deadlines? Why such an evident hierarchy in the punishments? Why the amplified powers of the state of siege and imposition of a deadline? Along the ease of approving constitutional changes, these last prerogatives were the third level of the chronology of exceptions: why? I will go back to the theme of the institutional framework in conclusion - soon - but it is surprising that the majority of the specialized literature does not attribute any importance to this, neither do Villa and Daniel, who propose the new periodizations. Well, concerning proposing periodizations, an accurate study of chronology is a crucial preliminary stage.

On July 22nd, 1964, as previously remembered, the National Congress approved the extension of Castelo Branco's term until March 15th, 1967, thus breaking the legality established by the institutional act. Being a measure approved by the parliament, it might not be considered an entirely authoritarian action, even if the Congress was purged due to the revoked mandates.

In the following year, a new "revolutionary outbreak" occurred due to AI-2, reopening the punishment seasons in October 27th, 1965, granting Castelo Branco new powers to revoke elected mandates and suppress political rights (precisely the most important exceptions) until March 15th, 1967, among other devices. It does not seem far-fetched to say Castelo Branco had full powers: the president could suppress the political powers of any citizen.

Thus, according to Villa's definition, it is possible to say that there was a dictatorship between the coup and June (or October) 1964. Then again between October 27th, 1965 and the end of Castelo's term on March 15th, 1967. Finally, the country would become a dictatorship after December 13th, 1968 with AI-5. Trying to understand Villa's proposal (that there was not a dictatorship in the initial stage), maybe we should suppose that, for him, in the first years there were, let us say, dictatorial spasms -- something that could not be properly and permanently characterized a dictatorship, reason which Villa also appeals to irony and uniqueness as key to Brazil. It would be a Brazilian dictatorship, as the title of his book.

Marshal Costa e Silva, were he still alive, would disagree with Villa. At the continuation of the National Security Council meeting on July 11th, 1968, on the 16th, trying to justify why he did not wish for a new institutional act at that moment, the marshal opened up to his ministers, in a confessional tone: "I must tell you, with sincere conviction, resulting from my days as a dictator in this country, that dictatorship will never be a solution to Brazil". That is, Costa e Silva had no doubts that he had been a dictator.^[41]

Daniel Aarão does not dwell on the consideration of the first moments after the coup when discussing periodization because he is more concerned about the end of the dictatorship but mentions the basic chronology after the Institutional Act. He does not question the different deadlines for the punishments and understands that "the deadline worked as a sort of commitment to restoring order and democracy." However, he is mistaken about the deadline, for he makes an inaccurate statement about the power to revoke mandates and suppresses political rights.^[42]

The closer to a conceptual definition of periodization he gets, Daniel weakly indicates the question while mistaking chronology for periodization: "chronological references are not objective, they are not there, before history is made, determining its ways [...] Chronology […] tends to order historical reflection, conditioning it in different directions, according to well-determined choices" (REIS, 2014, p. 127, emphasis in the original). To suppose that the ontological statute of dating is equivalent to the gnosiological construction of periodization is highly debatable. If we abdicate the realist philosophical hypothesis that something happened in 1964, in 1968, or who knows when everything becomes too difficult for historians. This has nothing to do with establishing periodizations, which surely do not result from adhesion to philosophical hypotheses of realism or idealism, being statements with pretense to truth remaining to be demonstrated. That is, establishing a chronology concerns the realist belief we share (or not, say the idealists) that, at some point, something happened: on December 13th, 1968 AI-5 was decreed. This belief results from our adhesion to the philosophical hypotheses of realism according to which - in grossly simplified manner - things happened, existed, and we are capable of knowing them. Another thing is the gnosiological periodization proposal, which might have a conventionalist or realistic epistemological motivation, but will always be a hypothesis to be demonstrated, unless it is a simple opinion. I deal with such theoretical questions in another work that can be consulted should the readership become interested (FICO, 2012).

As it is, Daniel argues that the dictatorship ended with the repealing of exception laws using the Constitutional Amend 11, on October 13th, 1978, which extinguished the institutional acts as of January 1st, 1979. "There was no longer a dictatorship, but there was yet to be a democracy." In a 2014 publication, he proposed that then a strange "authoritarian rule of law" appeared (REIS, 2014, p. 125). In a previous publication, Daniel had talked about a "precarious rule of law" (REIS, 2000, p. 69). His periodization is articulated with the denomination he proposes to the regime. "The chronological mark of 1985 extends a dense mantle of silence over the social and political - civil - basis of the dictatorship" (REIS, 2014, p. 128). That is, those who consider that the dictatorship only ended with the election of a civilian would be incapable of perceiving the importance of civilian support - even if the reason for such incapacity remains unclear. Daniel's reasoning is associated with his conviction that there would have been unconfessable interests behind the defense that the dictatorship ended in 1985. A "strange alliance" that would have legitimized the "silence about civilian participation in the coup process," an "interested memory," a "wide coalition of interests and wills," "true social, academic, and political front" (REIS, 2014, p. 127). The association between problematic theoretical foundation, limited empirical evidence, and perceiving political motivations as an apparent conspiracy does not favor Daniel Aarão Reis Filho's proposal.^[43]

Proposing a periodization of a given historical process is the noble attribution of History. Non-conventional proposals should not aspire to gather empirical evidence to support the attribute. Marco Antônio Villa's and Daniel Aarão Reis's postulates do not fulfill these and other requisites and are also weakened by the strong political connotation they contain. Insofar as they are weakly sustained, they seem mere personal opinion and, as such, easily overflow the academic environment to support a political reading according to which the regime that emerged with the 1964 coup was not very harmful to Brazil since its initial and final stages were not that strict.

5) Urgency and Tradition

The time of the military alternated between periods of "revolutionary" turbulence - understood as exception moments, of applying punishment - and phases of experience, of seeking relative institutional bonanza, that all military governments aimed at, except for general Médici, who "accepted the war," as seen. Thus understood, the temporality of generals related to the supposed urgency in cleaning the country - using cleaning operations, prisons, punishments - as well as to the allegedly democratic Brazilian tradition (thus the concern with institutionalization).

Thus, the regime's search for institutional normality supposed an incorporation of strict societal control mechanisms in the Constitution to make the institutional acts - which, despite their name, were not capable of institutionalize; to the contrary, they were exception and "revolutionary" - unnecessary, not allowing the normalization of the regime in the sense of its constitutional legality. Castelo Branco supposed to have gone back to constitutional normality when ended the validity of the most dangerous prerogatives of the Institutional Act, but soon came AI-2. For that reason, the end of his administration was the first and desperate attempt to institutionalize the regime, with the hurried approval of the new Constitution, of the Press Law, and of the National Security Law as measures to work as a safeguard to Costa e Silva, who would take office unattended of exception instruments. As seen, this apparatus would be considered inadequate in 1968.

These institutionalization attempts were opposed by episodes in which the regime was led to return to its "revolutionary sources" of 1964, moments that were also called "new revolutionary surge" or considered as the necessity to "reactivate the revolution" or to "restore the revolutionary principle." Naturally, the return to this matrix of exceptions caused the imposition by force of a new institutional act that would reopen the "revolutionary" punishment season, as happened in the first act. This occurred clearly with AI-2 and AI-5 regarding more important punishments, but the "revolutionary origins" were also activated with the AI-12, in September 1st, 1969. It was in fact a coup d'etat against vice-president Pedro Aleixo, impeded of taking presidential office when marshal Costa e Sliva fell ill, as well as AI-3 on a smaller scale, instituting indirect elections for governors on February 5th, 1966, and with AI-6 - that made alterations to the Supreme Court on February 1st, 1969. The institution of death and banishment penalties, by acts 13 and 14, on September 5th, 1969, motivated by the kidnapping of the North-American ambassador, attacked Brazilian juridical tradition, and are also an expression of those above "revolutionary" matrix.

In this sense, according to the military's understanding, since 1964 Brazil experienced phases of "revolutionary exception," as instruments they considered dictatorial and authoritarian, combined with periods of which the institutionalization of the regime was experienced or at least sought after. Between 1964 and 1985, while the generals were in power, it was always potentially possible to reactivate the revolution, because of the military - obviously - controlled the arms. For this reason, the period can be designated as a military dictatorship. Up until the inauguration of the first civilian president, the regime attributed itself the prerogative of a "revolutionary surge" which, in João Figueiredo's decadent phase, was synthesized in the general's threat of "calling Pires" -- that is, calling Minister of the Army General Walter Pires and thus solving problems with arms. President elect Tancredo Neves postponed a surgery because he feared non-constitutional solutions and died because of it. In fact, even if mitigated, the possibility of a backlash was present until the end of the military dictatorship and the inauguration of the first civilian president.

Marcos Napolitano approached the thematic of the institutional framework of the regime more closely. However, he stated that Castelo Branco's government was the "true institutional builder of the authoritarian regime. In his administration, four institutional acts were edited, as well as the Press Law and the new Constitution, sealing the principle of national security" (NAPOLITANO, 2016, p. 73). The problem with this statement is not in the factual error since Castelo only signed the first three institutional acts, but in supposing that the first president of the military regime was the author of its institutional framework.

If we understand as institutional design or institutional framework the normative and legal apparatus that establishes the structures that enable the State during the military regime, notably concerning the prominence of the Executive over the remaining powers and the citizens, in this case - diverging from Napolitano - it seems more accurate to claim that Costa e Silva was the main responsible for such a framework, which he established in the first institutional act, signed as Chief of the Supreme Command of the Revolution, having a brigadier and an admiral as supporters. Why do I affirm this?

As previously suggested, the regime debated the question of its legality regarding a dichotomy from the 1964 coup until 1985: normality versus exception. This problem is far from exclusive to the Brazilian dictatorship. In fact, it is secular for the Brazilian regime sought to equate such a dichotomy - which has its origin in the authoritarian claim to correct democracy using constitutional shortcuts - by employing a state of exception. As Giorgio Agamben suggests, it is possible to establish analogies between contemporary versions of the state of exception and the application of the Roman justitium as "space of law [...] anomie zone in which all judicial determinations [...] are deactivated":

This space of the law seems to be, in certain aspects, as essential to legal order that it should seek [...] to assure a relation with it, as if, to be founded, it needed to keep it [...] about an anomie [...] the space of law […] covers itself, to the legal order, of a decisive strategic relevance and that, in no way, can let it slip away. (AGAMBEN, 2004, p. 78-79)

Therefore, exception presupposes legality. Thus, the objective of the institutional acts was not the "legal reinforcement of the Executive Power" nor the search for an "authoritarian normalization" that would guarantee predictability to the system (NAPOLITANO, 2016, p. 79). The Institutional Act imposed by force of arms by Costa e Silva on April 9th, 1964 - against other institutional designs that, at that point, were under discussion at the National Congress - established those above dichotomous institutional framework. The following acts and constitutions were expressions of attempts at prevailing of the exceptional and institutional dimensions. The military dictatorship dealt with many constitutions or constitutional amends. The 1946 constitution, altered by AI-1; the 1967 constitution, by Castelo Branco; the one Costa e Silva requested to Pedro Aleixo in order to end AI-5, in 1969; the one bestowed by the military board in 1969, after Costa e Silva fell ill -- precisely the one coordinated by Pedro Aleixo, but without reference to the possibility of extinguishing the act. Finally, the alterations established by the Constitutional Amend 11 in 1978, by Ernesto Geisel. As is known, only Ernesto Geisel's institutionalization project was successful.

In the aforementioned National Security Council meeting, in which marshal Costa e Silva admitted to having acted as dictatorship at the beginning of the regime, he said that "revolution was not carried out to oppress the people. On the contrary, it was carried out as a counter-revolution, precisely to avoid the oppression of the people [...] If we wanted to beat tyranny - said the president - we could not establish another tyranny". A longer passage by the marshal is worthwhile transcribing:

But we were able, with God's grace and men's comprehension, to beat that first stage by institutionalizing the revolution [...] What we wished for was reestablishing order and the Constitution. Evidently, we could not attain these objectives simply by changing the men in power. There should have been, as there was, an act to allow the revolution to [...] generate its rights [...] with the elimination of the means of disturbance [...] there was a truce and this truce, afterward, as taken as in the sense of condescension and tolerance [...] thus act number two. Amazingly, the act number two was not edited due to the actions of those who had their mandates revoked. It resulted from the actions of the good revolutionaries, at first [...] What I want to highlight, however, is that, from the very first moment of the revolution, there was a preoccupation, which still characterizes the revolutionary process, with returning to a fully democratic regime.^[44]

Costa e Silva believed to be leading a fully constitutional government, since AI-2 lost its validity on the eve of his inauguration. When the minister of the Navy supported the thesis that the 1968 demonstrations could have been combated without the aid of exception instruments, the marshal welcomed the thesis with visible satisfaction and said that they could count on the "little book," as Dutra would say.^[45] When the meeting ended, and the elaboration of a note to press was discussed, Costa e Silva determined that the institutional acts should not be mentioned not even as a threat, because "should it come to that, it will be a huge sacrifice and with the regret of an old revolutionary, for it will be a setback to the ongoing process".^[46]

The act would end up coming five months later, so that the marshal, seen as an uneducated authoritarian figure by those who mocked him, was responsible for two of the main exception designs of the regime (AI-1 and AI-5). However, Costa e Silva did not give up on the institutionalization process. He determined that vice-president Pedro Aleixo made a new Constitution to allow the gradual end of the institutional acts. Pedro Aleixo handed in 300 pages on June 24th, 1969. Costa e Silva read and annotated them by the beginning of July. Then, between July 14th and 17th 1969, the marshal personally coordinated the commission he summoned to discuss the proposal. His ministers Rondon Pacheco (Civil Office) and Helio Beltrão (Planning), Supreme Court minister Temístocles Cavalcanti, former Supreme Court minister Carlos Medeiros, and the dean of Universidade de Sao Paulo, Miguel Reale, participated in this commission. Having concluded the discussions, the project established in its article 182 that "the president of the Republic, when considering the national interest, will cease, by means of a decree, the validity of any or all constant devices of the Institutional Act n. 5, of December 13th, 1968, and of other later approved acts" (ALEIXO, 2002). The project was also founded on a safeguard: amplifying the president of the Republic's powers to decree State of Siege, which could then last 180 days (Castelo's 1967 Constitution had increased the 30 days in the 1946 Constitution to 60).

Costa e Silva planned to re-open the National Congress (closed since AI-5) and enact the constitutional amend on September 1st, 1969. The military ministers did not approve one or the other but could not oppose the leader of the 1964 movement who had signed the institutional act and AI-5. Three days before the scheduled date, the marshal suffered a thrombosis that would remove him from office permanently. Subsequently, his military ministers carried out a coup d'etat, preventing vice-president Pedro Aleixo of taking presidential office, modified essential points of the project, and promulgated the Constitution with alterations. The new wording of article 182 became: "the Institutional Act n. 5 of December 13th, 1968 is still valid, as well as other later approved institutional acts". Such details are ignored or obscured by specialized literature. A simplified narrative still prevails, portraying Costa e Silva as leader of the so-called hard line and responsible for the AI-5 still - which would have been, according to a paradoxical and ironic cliché, the "coup inside the coup", when in fact AI-5 was simply another manifestation of the dynamic between the institutionalizing and the exception tendencies. AI-5 did not change the nature of the regime. After 1964, a new coup indeed took place in 1969, against Pedro Aleixo.

Years later, Ernesto Geisel would face similar problems, caused by "good revolutionaries," who in his case were "sincere but radical revolutionaries," ^[47]an expression he used in a speech delivered before the Arena leaders, on December 1st, 1977, at Palacio da Alvorada. It was in this speech that Geisel announced, in definitive terms, his project of "putting an end to exception laws [...] replacing them by adequate constitutional safeguards, allowing [...] the maintenance and the better functioning of the democratic regime and order" (BRASIL. PRESIDENCY OF THE REPUBLIC, 1996, p. 340-341). Such safeguards would be the State of Emergency and the Emergency Measures, included in the Constitutional Amend n. 13, of October 13th, 1978, which extinguished the exception acts. The State of Emergency allowed the application of all State of Siege measures by the President of the Republic, without needing approval from the National Congress whenever the powers or institutions were "severely threatened or harmed by subversion factors." According to jurist Enrique Ricardo Lewandowski,

In the hypothesis of decreeing or extending the State of Emergency, the presidential decision is merely communicated to the Congress, five days after being taken by the head of government. In this case, the legislature is no more than a spectator of the application of one of the most violent medicines to preserve the existing institutions in the constitutional order [...] the president has vast autonomy to decree State of Emergency since he is in no way restrained by the other powers of the Republic [...] Consulting the Constitutional Council is merely a formality. (LEWANDOWSKI, 1984, p. 158)

Such safeguards would be valid until the 1988 Constitution. Emergency measures were utilized in the Brasilia region in October 1983 (due to polling on wage policy), and in April 1984 to pressure the National Congress to reject the amendment for direct elections. In this occasion, President João Figueiredo said that, should the amendment be approved, the risk of a coup would be real. Such information was given to allied congressman Francisco Salles, who had a meeting with the general.^[48] The possibility of using State of Emergency and the Emergency Measures makes Daniel Aarão Reis Filho's and Marco Antonio Villa's periodization proposals even more debatable towards the end of the dictatorship.

This was the background related to the institutional framework of the regime that inspired Ernesto Geisel to design the project of a "slow, gradual, and safe distension." As argued by Suzeley Kalil Mathias in a master's thesis defended in 1992 and published in 1995, Geisel's project did not come out of nowhere, since it contained demands from both the opposing and the supporting parties, "solid basis upon which he could lay his project." But there was also resistance, in case "not enough guarantees were given to the sectors committed to the dictatorship and political repression that they would not be judged" (MATHIAS, 1995, p. 62, 65, and 68). Above all, there were institutionalization attempts based on the Constitutionalization since marshal Castelo Branco's administration, using safeguards that would make the State of Siege and similar devices even stricter. Such safeguards were always thought of as replacements for social control mechanisms, present in the institutional acts. That is precisely what Geisel did. The difference is that, contrary to Castelo Branco and Costa e Silva, he succeeded. Médici simply did not try, even though he established passing on the government to a civilian, in case the "armed struggle" was completely defeated. According to his son, he would opt for an active general if the confrontation were serious, and to an army reserve general "if things were halfway through" - the alternative he ended up choosing (MÉDICI, 1995, p. 31). Geisel was chosen and, therefore, his distention project did not result from his psychological profile (as suggested in Gaspari's work) but rather from the historical accumulation of such experiences.

Several analysts suggest that Geisel's initial project was modified due to societal pressure.^[49] Like me, Suzeley Kalil Mathias supports the existence of a Geisel project. However, she judges that amnesty and party reform resulted from the government's need to "negotiate reforms and extend limits" (MATHIAS, 1995, p. 143). However, the dissolution of political parties was under discussion by Geisel and Golbery at least since 1975, as much as the end of the AI-5 and amnesty (GASPARI, 2004, p. 89-91). In other words, Geisel's project was designed since the beginning of his government and foresaw all these stages, with self-amnesty, weakening the Brazilian Democratic Movement (MDB), ending bipartisanship, and choosing the new military president by indirect vote the most important ones. Thus, despite the campaign, amnesty was not "broad, general, and unrestricted" and forgave the torturers. New parties were created. The Diretas Já (Direct Elections Now) campaign did not stop the electoral college from acting once again. If Democratic resistance did not speed up the "opening" project, was it unimportant? Those are different things: the social dynamics of the time was important for the configuration of democracy in our country today. The 1988 Constitution was marked by demands that mobilized social manifestations by the end of the dictatorship. But the exit of the military regime was controlled by the military. Francisco Carlos Teixeira da Silva called attention to the fact that Geisel guaranteed that "there was no project." Curiously, after this statement, the former president described the general principles of the distention and appointed the hard line as an actor alongside the opposition - that is, described his project (SILVA, 2003, p. 264).

It was also Francisco Carlos who distinguished an important aspect: at the end of the process, Paulo Maluf's insistence in running for president "could, dangerously, shatter Golbery's/Geisel's/Figueiredo's initial architecture, throwing the country in financial and economic trouble, to a transition by collapse." Francisco Carlos, in this passage, is referencing general Figueiredo's immobility in the final phase of his term, which prevented him from coordinating his succession, and Paulo Maluf's offensive to become the candidate of the supporting party -- even if he had to risk the very scheme that supported the regime to do so. Political parties, then, took the lead in the process "to avoid backlashes or ruptures" and, indeed, organized Maluf's defeat and the exit of dictatorship. Francisco Carlos affirms that this "was no longer Geisel and Golbery's opening." To me, it seems that Geisel and Golbery's "opening" was being carried out despite the lack of leadership from Figueiredo and the turbulence caused by Maluf (SILVA, 2003, p. 273-275. Emphasis in the original). Therefore, the Democratic Alliance and the New Republic were the crownings of a slow, gradual, and safe distension.

I do not think it is necessary, to sum up, in conclusion, what was previously said. I would simply like to reaffirm, to conclude, my conviction that the theoretical-conceptual and historiographical debate of these important questions are only possible due to the high quality of the production of the authors here mentioned.

References

AGAMBEN, Giorgio. Estado de exceção. Tradução de Iraci D. Poleti. São Paulo: Boitempo, 2004.

ALEIXO, José Carlos Brandi. Pedro Aleixo, organizador e relator da reforma constitucional. In: SENADO FEDERAL. A constituição que não foi: história da Emenda Constitucional n. 1, de 1969. Brasília; Senado Federal, 2002.

ALMEIDA, Guilherme Bacha de. O Conselho de Segurança Nacional e a ditadura (1964-1969). In: MOTTA, Rodrigo Patto Sá (Org.). Ditaduras militares: Brasil, Argentina, Chile e Uruguai. Belo Horizonte: UFMG, 2015.

ARAUJO, Maria Paula. A luta democrática contra o regime militar na década de 1970. In: REIS, Daniel Aarão; RIDENTI, Marcelo; MOTTA, Rodrigo Patto Sá (Orgs.). O golpe e a ditadura militar: 40 anos depois (1964-2004). Bauru: Edusc, 2004.

BERSTEIN, Serge. Culturas políticas e historiografia. In: AZEVEDO, Cecília e outros. Cultura política, memória e historiografia. Rio de Janeiro: FGV, 2009.

BRASIL. PRESIDÊNCIA DA REPÚBLICA. Governos da República: 15.11.1889 a 15.03.1996. Presidência da República, Casa Civil, Biblioteca. 3ª ed. rev. e atual. Brasília, 1996.

CARDOSO, Ciro Flamarion Santana. Epistemologia pós-moderna, texto e conhecimento: a visão de um historiador. Diálogos. Maringá, v. 3, n. 3, p. 1-28, 1999.

CHARTIER, Roger. Le monde comme représentation. Annales E.S.C. Paris, nov./dec./ 1989 n. 6 p. 1505-1520.

CARDOSO, Fernando Henrique. O modelo político brasileiro e outros ensaios. São Paulo: Difel, 1972.

CHAUÍ, Marilena. Apontamentos para uma crítica da Ação Integralista Brasileira. In: CHAUÍ, Marilena; FRANCO, Maria Sylvia de Carvalho. Ideologia e mobilização popular. Rio de Janeiro: Paz e Terra, 1978.

CODATO, Adriano Nervo. O golpe de 1964 e o regime de 1968: aspectos conjunturais e variáveis históricas. História: Questões & Debates. Curitiba, n. 40, p. 11-36, 2004.

D’ARAUJO, Maria Celina. Sindicatos, carisma e poder: o PTB de 1945-65. Rio de Janeiro: FGV, 1996.

DANIEL, Herbert. Passagem para o próximo sonho. Rio de Janeiro: Codecri, 1982.

DIAS, Luiz Antonio. Entre letras e números: uma análise do jornal Folha de S. Paulo e de pesquisas de opinião do Ibope (1963-1964). In: FERNANDES, Carla M.; CHAGAS, Genira. (Orgs.). Mídia e governos autoritários: 60 anos do suicídio de Getúlio Vargas e 50 anos do golpe civil-militar no Brasil. João Pessoa: Idea, 2014.

FERREIRA, Jorge. O governo Goulart e o golpe civil-militar de 1964. In: FERREIRA, Jorge; DELGADO, Lucília de Almeida Neves (Orgs.). O Brasil republicano: o tempo da experiência democrática – da democratização de 1945 ao golpe civil-militar de 1964 (Livro 3). Rio de Janeiro: Civilização Brasileira, 2003.

FERREIRA, Jorge. Entre o comício e a mensagem: o presidente Goulart, as esquerdas e a crise política de março de 1964. In: AZEVEDO, Cecília e outros. Cultura política, memória e historiografia. Rio de Janeiro: FGV, 2009.

FICO, Carlos. Reinventando o otimismo: ditadura, propaganda e imaginário social no Brasil. Rio de Janeiro: FGV, 1997.

FICO, Carlos. Como eles agiam. Os subterrâneos da ditadura militar: espionagem e polícia política. Rio de Janeiro: Record, 2001.

FICO, Carlos. Espionagem, polícia política, censura e propaganda: os pilares básicos da repressão. In Ferreira, Jorge, Neves, Lucilia de Almeida (Orgs.). O Brasil republicano. Livro 4: O tempo da ditadura – regime militar e movimentos sociais em fins do século XX. Rio de Janeiro: Record, 2003.

FICO, Carlos. Além do golpe: versões e controvérsias sobre 1964 e a ditadura militar. Rio de Janeiro: Record, 2004.

FICO, Carlos. O grande irmão. Da Operação Brother Sam aos anos de chumbo: o governo dos Estados Unidos e a ditadura militar brasileira. Rio de Janeiro: Civilização Brasileira, 2008.

FICO, Carlos. A ditadura documentada. Acervo. Rio de Janeiro, v. 21, n. 2, p. 67-78, jul./dez. 2008a.

FICO, Carlos. 1968: o ano que terminou mal. In: FICO, Carlos; ARAÚJO, Maria Paula (Orgs.). 1968, 40 anos depois: história e memória. Rio de Janeiro: 7Letras, 2009.

FICO, Carlos. Les régimes autoritaires au Brésil durant l’ére républicaine. Brésil(s). Sciences humaines et sociales. Paris, v. 1, Ouverture, p. 30-40, 2011.

FICO, Carlos. História que temos vivido. In: VARELLA, Flávia e outros (Orgs.). Tempo presente & usos do passado. Rio de Janeiro: FGV, 2012.

FICO, Carlos. Violência, trauma e frustração no Brasil e na Argentina: o papel do historiador. Topoi. Revista de História. Rio de Janeiro, v. 14, n. 27, p. 239-284, jul./dez. 2013.

FICO, Carlos. História do Brasil contemporâneo: da morte de Vargas aos dias atuais. São Paulo: Contexto, 2015.

FIGUEIREDO, Argelina Cheibub. Democracia ou Reformas? Alternativas democráticas à crise política: 1961-1964. Tradução de Carlos Roberto Aguiar. Rio de Janeiro: Paz e Terra, 1993.

FONTES, Virgínia. Nas lutas. In: MELO, Demian Bezerra de (Org.) A miséria da historiografia: uma crítica ao revisionismo contemporâneo. Rio de Janeiro: Consequência, 2014.

GARCIA, Miliandre. “Ou vocês mudam ou acabam”: teatro e censura na ditadura militar (1964-1985). Tese de doutorado apresentada ao Programa de Pós-graduação em História Social da UFRJ. Rio de Janeiro, 2008.

GASPARI, Elio. A ditadura encurralada. São Paulo: Companhia das Letras, 2004.

GORENDER, Jacob. Combate nas trevas. A esquerda brasileira: das ilusões perdidas à luta armada. 2. ed. São Paulo: Ática, 1987.

GORENDER, Jacob. A escravidão reabilitada. LPH – Revista de História. Mariana, v. 3, n. 1, p. 245-266, 1992.

HEMPEL, Carl G. The Function of General Laws in History. The Journal of Philosophy. New York, v. 39, n. 2, p. 35-48, Jan. 1942.

HOBSBAWM, Eric. Não basta a história da identidade. In: __________. Sobre história: ensaios. Tradução de Cid Knipel Moreira. São Paulo: Companhia das Letras, 1998.

LACAPRA, Dominick. Writing History, Writing Trauma. Baltimore: The Johns Hopkins University Press, 2001.

LARA, Silvia Hunold. Escravidão no Brasil: balanço historiográfico. LPH – Revista de História. Mariana, v. 3, n. 1, p. 215-244, 1992.

LEWANDOWSKI, Enrique Ricardo. As salvaguardas do Estado no Brasil. Revista da Faculdade de Direito. São Paulo, v. 79, p. 140-160, jan. 1984.

MARTINS FILHO, João Roberto. Movimento estudantil e ditadura militar: 1964-1968. Campinas: Papirus, 1987.

MARTINS, Franklin. Prefácio. In: GURGEL, Antonio de Padua. A rebelião dos estudantes: Brasília, 1968. Brasília: UnB, 2002.

MARTINS, William de Souza Nunes. Produzindo no escuro: políticas para a indústria cinematográfica brasileira e o papel da censura (1964-1988). 2009. Tese (doutorado em História Social ) – UFRJ, Programa de Pós-graduação em História Social, Rio de Janeiro, 2009.

MATHIAS, Suzeley Kalil. Distensão no Brasil: o projeto militar (1973-1979). Campinas: Papirus, 1995.

MATTOS, Marcelo Badaró. O governo João Goulart: novos rumos da produção historiográfica. Revista Brasileira de História. São Paulo, v. 28, n. 55, p. 245-263, 2008.

MATTOS, Marcelo Badaró. As bases teóricas do revisionismo: o culturalismo e a historiografia brasileira contemporânea. In: MELO, Demian Bezerra de (Org.) A miséria da historiografia: uma crítica ao revisionismo contemporâneo. Rio de Janeiro: Consequência, 2014.

MÉDICI, Roberto Nogueira. Médici, o depoimento. Rio de Janeiro: Mauad. 1995.

MELO, Demian Bezerra de (Org.) A miséria da historiografia: uma crítica ao revisionismo contemporâneo. Rio de Janeiro: Consequência, 2014.

MELO, Demian Bezerra de. A miséria da historiografia. Outubro. n. 14, p. 111-130, 2º. sem. 2006.

MELO, Demian Bezerra de. O golpe de 1964 e meio século de controvérsias: o estado atual da questão. In: MELO, Demian Bezerra de (Org.) A miséria da historiografia: uma crítica ao revisionismo contemporâneo. Rio de Janeiro: Consequência, 2014.

MOORE JR. Barrington. As origens sociais da ditadura e da democracia. Tradução de Maria Ludovina F. Couto. São Paulo: Martins Fontes, 1983.

MOTTA, Rodrigo Patto Sá. As universidades e o regime militar: cultura política brasileira e modernização autoritária. Rio de Janeiro: Zahar, 2014.

NAGEL, Ernest. Some issues in the Logic of the Historical Analysis. The Scientific Monthly. New York, v. 74, n. 3, p. 162-169, March 1952.

NAPOLITANO, Marcos. A “resistência cultural” durante o regime militar brasileiro: um novo olhar historiográfico. In: MOTTA, Rodrigo Patto Sá (Org.). Ditaduras militares: Brasil, Argentina, Chile e Uruguai. Belo Horizonte: UFMG, 2015.

NAPOLITANO, Marcos. 1964: história do regime militar brasileiro. São Paulo: Contexto. 2016.

OLIVEIRA, Eliézer Rizzo de. As Forças Armadas: política e ideologia no Brasil (1964-1969). Petrópolis: Vozes, 1976.

OLIVEIRA, Gil Vicente Vaz. Fotojornalismo subversivo: 1968 revisto pelas lentes do Correio da Manhã. Acervo, Rio de Janeiro, v. 11, nº 1-2, p. 117-136, jan./dez., 1998.

PALMEIRA, Vladimir. Vladimir Palmeira [entrevista concedida por]. In: REIS FILHO, Daniel Aarão; MORAES, Pedro de. 1968: a paixão de uma utopia. 3. ed. Rio de Janeiro: FGV, 2008.

REIS, Daniel Aarão. Ditadura militar, esquerdas e sociedade. Rio de Janeiro: Jorge Zahar, 2000.

REIS, Daniel Aarão. Ditadura e sociedade: as reconstruções da memória. In: REIS, Daniel Aarão; RIDENTI, Marcelo; MOTTA, Rodrigo Patto Sá (Orgs.). O golpe e a ditadura militar: 40 anos depois (1964-2004). Bauru: Edusc, 2004.

REIS, Daniel Aarão. Trinta anos depois. In: REIS FILHO, Daniel Aarão; MORAES, Pedro de. 1968: a paixão de uma utopia. 3. ed. Rio de Janeiro: FGV, 2008.

REIS, Daniel Aarão. Ditadura, anistia e reconciliação. Estudos Históricos. Rio de Janeiro, v. 23, n. 45, p. 171-186, jan./jun. 2010.

REIS, Daniel Aarão. Ditadura e democracia no Brasil: do golpe de 1964 à Constituição de 1988. Rio de Janeiro: Zahar, 2014

REIS, Daniel Aarão; ROLLAND, Denis. Apresentação. In: REIS, Daniel Aarão; ROLLAND, Denis (Orgs.). Modernidades alternativas. Rio de Janeiro: FGV, 2008.

REIS FILHO, Daniel Aarão. A revolução faltou ao encontro: os comunistas no Brasil. São Paulo: Brasiliense, 1980.

REIS FILHO, Daniel Aarão e outros. Versões e ficções: o sequestro da história. São Paulo: Perseu Abramo, 1997.

REIS FILHO, Daniel Aarão. 1968, o curto ano de todos os desejos. Tempo Social. São Paulo, v. 10, n. 2, p. 25-35, out. 1998.

REIS FILHO, Daniel Aarão. Ditadura no Brasil: entre memória e história. In: MOTTA, Rodrigo Patto Sá (Org.). Ditaduras militares: Brasil, Argentina, Chile e Uruguai. Belo Horizonte: UFMG, 2015.

REIS FILHO, Daniel Aarão; MORAES, Pedro de. 1968: a paixão de uma utopia. 3. ed. Rio de Janeiro: FGV, 2008.

RIDENTI, Marcelo. O fantasma da revolução brasileira. São Paulo: UNESP, 1993.

RIDENTI, Marcelo. Que história é essa? In: REIS FILHO, Daniel Aarão e outros. Versões e ficções: o sequestro da história. São Paulo: Perseu Abramo, 1997.

RIDENTI, Marcelo. Cultura e política: os anos 1960-1970 e sua herança. In: FERREIRA, Jorge; DELGADO, Lucília de Almeida Neves (Orgs.). O Brasil republicano: o tempo da ditadura – regime militar e movimentos sociais em fins do século XX (Livro 4). Rio de Janeiro: Civilização Brasileira, 2003.

RIDENTI, Marcelo. Canetas e fuzis: intelectuais e artistas brasileiros nos anos 1960/70. In: REIS, Daniel Aarão; ROLLAND, Denis (Orgs.). Modernidades alternativas. Rio de Janeiro: FGV, 2008.

ROLLEMBERG, Denise. O apoio de Cuba à luta armada no Brasil: o treinamento guerrilheiro. Rio de janeiro: Mauad, 2001.

ROLLEMBERG, Denise. Esquerdas revolucionárias e luta armada. In: FERREIRA, Jorge; DELGADO, Lucília de Almeida Neves (Orgs.). O Brasil republicano: o tempo da ditadura – regime militar e movimentos sociais em fins do século XX (Livro 4). Rio de Janeiro: Civilização Brasileira, 2003.

ROLLEMBERG, Denise. Memória, opinião e cultura política. A Ordem dos Advogados do Brasil sob a ditadura (1964-1974). In: REIS, Daniel Aarão; ROLLAND, Denis. (Orgs.). Modernidades alternativas. Rio de Janeiro: Editora da Fundação Getúlio Vargas, 2008, p. 57-96.

ROLLEMBERG, Denise. As trincheiras da memória. A Associação Brasileira de Imprensa e a ditadura (1964-1974). In: ROLLEMBERG, Denise; QUADRAT, Samantha Viz. (Orgs.). A construção social dos regimes autoritários. Legitimidade, consenso e consentimento no Século XX: Brasil e América Latina. Rio de Janeiro: Civilização Brasileira, 2010, vol. 2, p. 97-144.

ROLLEMBERG, Denise; QUADRAT, Samantha Viz. (Orgs.). A construção social dos regimes autoritários. Legitimidade, consenso e consentimento no Século XX: Brasil e América Latina. Rio de Janeiro: Civilização Brasileira, 2010, vol. 2.

SAES, Décio. Classe média e política no Brasil (1930-1964). In: FAUSTO, Boris (Dir.). História geral da civilização brasileira. Tomo III: O Brasil republicano, 3º volume: Sociedade e política (1930-1964). São Paulo: Difel, 1981.

SCARTEZINI, Antonio Carlos. Segredos de Medici. São Paulo: Marco Zero, 1985.

SCHMITTER, Philippe C. Interest conflict and political change in Brazil. California: Stanford University Press, 1971.

SCHNEIDER, Ronald M. The political system of Brazil: emergence of a “modernizing” authoritarian regime, 1964-1970. New York: Columbia University Press, 1971.

SERRA, Jose´. Cinquenta anos esta noite: o golpe, a ditadura e o exílio. Rio de Janeiro: Record, 2014.

SILVA, Francisco Carlos Teixeira da. Crise da ditadura militar e o processo de abertura política no Brasil, 1974-1985. In: FERREIRA, Jorge; DELGADO, Lucília de Almeida Neves (Orgs.). O Brasil republicano: o tempo da ditadura – regime militar e movimentos sociais em fins do século XX (Livro 4). Rio de Janeiro: Civilização Brasileira, 2003.

SILVA, Izabel Priscila Pimentel da. Os filhos rebeldes de um velho camarada: a dissidência comunista da Guanabara (1964-1969). 2009. Dissertação (Mestrado em História) - UFF, Programa de Pós-graduação em História, Niterói, 2009.

SIRKIS, Alfredo. Os carbonários. Rio de Janeiro: Record, 1998.

SOUZA, Jessé. A tolice da inteligência brasileira. São Paulo: LeYa, 2015.

STACCHINI, José. Março 64: mobilização da audácia. São Paulo: Companhia Editora Nacional, 1965.

STEPAN, Alfred. The military in politics: changing patterns in Brazil. Princeton University Press, 1971.

TOLEDO, Caio N. de. 1964: golpismo e democracia. As falácias do revisionismo. Crítica Marxista. Campinas, n. 19, p. 27-48, 2004.

VIEIRA, Fabrício de Assis C. Vieira; HOLLAND, Márcio Holland. Crescimento econômico secular no Brasil, modelo de Thirlwall e termos de troca. Economia e Sociedade. Campinas, v. 17, n. 2 (33), p. 17-46, ago. 2008.

VILLA, Marco Antonio. Jango: um perfil (1945-1964). São Paulo: Globo, 2004.

VILLA, Marco Antonio. Ditadura à brasileira – 1964-1985: a democracia golpeada à esquerda e à direita. São Paulo: LeYa, 2014.

Notes