Stability of Patriotism 1 Running head: STABILITY OF PATRIOTISM The Stability of Patriotism in the Face of Variation in National Laws

Patriotism has been described by several authors as a rigid and stable construct. It is argued that this might lead to a deterministic conception of citizenship, in which citizens of a certain country are supposed to react in a pre-determined way, as far as factors affecting attachment to the country are concerned. It is proposed that such a view conceals the fact that people’s relations to the country, even affective relations such as attachment, are political relations, and therefore, should be affected by the political activity of the State. This view received some support from an experimental study in which French citizens expressed their level of blind and constructive patriotism (Schatz & Staub, 1997) after having been exposed to a bogus national law that favored either immigrants (pro-outgroup only), nationals (proingroup only), or both. Blind patriotism appeared stable across contexts, whereas constructive patriotism varied. Implications for the study of patriotism and citizenship are discussed. Stability of Patriotism 3 This is no time for ‘my country right or wrong’ Remember what that brought Lou Reed Introduction How many things do people do on behalf of their country? They cry when listening to the national anthem, they accept to fight a war, they help their neighbors, they recycle their waste, and so on. All these actions, and many others, express a personal attachment towards one’s own country and are commonly referred to as patriotism. However, is patriotism a unidimensional, stable, concept that unequivocally defines what one should and should not do to support the country? Or, if patriotism is a relative concept –as it will be argued—, what makes it vary? These questions address the problem of articulating patriotism within the larger framework of citizenship—the civil and political rights and duties that are a part of the social life for each member of a democratic society. As Janowitz (1983) pointed out, the sociopolitical conception of citizenship has to be examined in terms of patriotism, ideology, and civil consciousness. For him, citizenship implies a kind of obedience to society but also a certain support from the citizens of the social values and the political institutions. As argued above, patriotism can take a large number of forms, and discussing the problem of its phenomenology—although important—is not the aim of the present article. However, it is possible to provide a core definition of this concept: “Patriotism is the attachment by group members to their group and the land in which it resides” (Bar-Tal & Staub, 1997b, p. 2). Accordingly, patriotism is supposed to be a fundamental element of individual and group life, and it is believed to fulfill such basic needs as the needs for security, for a positive identity, for effectiveness and control, for a positive connection to other people, and for comprehension of reality (Staub, 1989; 1996). Although patriotism is a very ancient phenomenon, it has attracted a great deal of attention in recent writings in the area of social psychology. In particular, an edited book (Bar-Tal & Staub, 1997a) and a Stability of Patriotism 4 recent debate in the British Journal of Social Psychology in 2001 have pointed to a number of critical features of patriotism that makes it a controversial issue when discussing citizenship, and in particular the relation between the individual and the country’s values (e.g. Mummendey, Klink and Brown, 2001; Condor, 2001; Hopkins, 2001). Such controversies include the relation between patriotism and nationalism (e.g. Kosterman and Feshbach,1989; Blank, Schmidt, and Westle, 2001), the “good” or “bad” nature of these concepts, (e.g. Janowitz, 1983; Connor, 1993), the conditions under which patriotism and/or nationalism induce out-group derogation (Blank et al., 2001; Mummendey, Klink and Brown, 2001), and the stable and fundamental nature of patriotism. The present article does not address all of these problems, but rather focuses on the latter: the stability of patriotism. This is, in our view, a problem for citizenship. More specifically, the stability of patriotism is critical when dealing with the ability of the democratic alternation (the alternation in power of parties that hold different values and goals, typical of democratic countries) to induce different representations of the country, as well as different forms of attachment to it. Indeed, if patriotism is a stable feature for the members of a country, for example if it is acquired through the exposure to national history and iconography (Reykowski, 1997), then it should show little sensitivity to variations in the values expressed by the different groups, parties and coalitions that alternate in power in a democracy. If this is true—and we will argue against it—it may be a problem, because it would imply that attachment to and representation of the country depend only upon long-term historical factors and not upon the political activity of the State, such as national decisions, the orientation of a specific government (that in most democratic countries is in power “only” for four-five years), and the discourses of politicians. The stability of patriotism The recent literature on the roots of patriotism seems to argue for a stable “nature” of patriotism. For instance, Johnson (1997) argues for an evolutionary approach, considering Stability of Patriotism 5 patriotism as an inborn predisposition that applies to all social systems and is insensitive to short-term incentives. In this view, patriotism is considered as a basic and fundamental force in “human nature”, anchored in such evolutionary mechanisms as nepotism and reciprocity (Buss, 1987). In the same vein, Feshbach and Sakano (1997) argue that attachment to the country is a fundamental process, although they find its origin in the generalization of affect from parents to nation: because the nation fulfills for the adult the same needs as the ones that parents fulfill for the child, attachment to the country is understood as a deep loyalty to, literally, the “fatherland”. Indeed, these authors found empirical support for this assumption by showing that patriotism is correlated with attachment to the father. It is interesting to note that this position is quite similar in structure to the one defended by Adorno, FrenkelBrunswik, Levinson and Sanford in 1950, when authoritarianism was considered a stable personality trait, rooted in the parents’ education. In spite of the abundant criticism that this approach has generated (see for example Brown, 1996), the temptation to consider social and political life as definitively determined by early interaction with parents is still widespread. Historical approaches to patriotism (e.g. Ben-Amos, 1997) also tend to consider the development of patriotism as based on non-contingent factors. In this view, patriotism is determined by such factors as the teaching of national history in schools, or the building of monuments for past heroes; it follows that patriotism is bound to remain a quite rigid construct that can only evolve over several generations. In sum, there is a significant tendency in the literature to conceive patriotism as a stable construct. This might lead to a deterministic conception of citizenship, in which citizens of a certain country are supposed to react in a pre-determined way, as far as factors affecting attachment to the country (e.g. aggression, relations to immigrants) are concerned. The present article proposes that such a view conceals the fact that people’s relations to the country, even affective relations such as attachment, are political relations, and therefore Stability of Patriotism 6 should be affected by the political activity of the State. This was proposed by Branson (2002): “The citizen who truly loves this country must first understand and be freely committed to its ideals—the values and principles to which it is committed. The citizen who truly loves this Republic must also be willing to take part in safeguarding those values and principles and in helping to effect the more complete realization of this nation’s ideals. In short, the citizen must be both ‘enlightened’ and ‘engaged’”. In other words, patriotism should be partly related to the political state of affairs. This view is consistent with Habermas’ concept of “constitutional patriotism” (1996), a form of attachment to a constitutional order, i.e. a particular national interpretation of more general constitutional principles, such as human rights. The idea that patriotism can be related to high-order political considerations is also pointed out by the notion of “enlightened political socialization” (Habermas, 1996), which relates to the fact that culture, and cultural byproducts such as patriotism, are not reproduced dogmatically through blind attachment to tradition (Hendley, 1998), but rather through active political relationships. Moreover, Habermas’ ideas are in line with Sandel (1982) and Taylor (1989) who propose that the national sense of community stems from a common appreciation of the support for liberal (in the classic sense) political principles. Thus, even if patriotism can be considered as an attachment to the nation based on an affective dimension, it also appears important to consider the influence of political factors. This, in a way, is in line with Habermas’ (1992) distinction between the ethnic-cultural and the democratic sides of citizenship, also called Ethnos and Demos dimensions of national citizenship (see also Sanchez-Mazas, van Humskerken, and Casini, 2003). In sum, the above work concurs with the argument in favor of the variability of patriotism, at least on the basis of relevant political factors. Based on this reasoning, the first, general, hypothesis is that patriotism should be influenced by the political decisions taken by the State. In the research that will be presented, Stability of Patriotism 7 French nationals were exposed to ex


Introduction
How many things do people do on behalf of their country?They cry when listening to the national anthem, they accept to fight a war, they help their neighbors, they recycle their waste, and so on.All these actions, and many others, express a personal attachment towards one's own country and are commonly referred to as patriotism.However, is patriotism a unidimensional, stable, concept that unequivocally defines what one should and should not do to support the country?Or, if patriotism is a relative concept -as it will be argued-, what makes it vary?These questions address the problem of articulating patriotism within the larger framework of citizenship-the civil and political rights and duties that are a part of the social life for each member of a democratic society.As Janowitz (1983) pointed out, the sociopolitical conception of citizenship has to be examined in terms of patriotism, ideology, and civil consciousness.For him, citizenship implies a kind of obedience to society but also a certain support from the citizens of the social values and the political institutions.
As argued above, patriotism can take a large number of forms, and discussing the problem of its phenomenology-although important-is not the aim of the present article.
However, it is possible to provide a core definition of this concept: "Patriotism is the attachment by group members to their group and the land in which it resides" (Bar-Tal & Staub, 1997b, p. 2).Accordingly, patriotism is supposed to be a fundamental element of individual and group life, and it is believed to fulfill such basic needs as the needs for security, for a positive identity, for effectiveness and control, for a positive connection to other people, and for comprehension of reality (Staub, 1989;1996).Although patriotism is a very ancient phenomenon, it has attracted a great deal of attention in recent writings in the area of social psychology.In particular, an edited book (Bar-Tal & Staub, 1997a) and a recent debate in the British Journal of Social Psychology in 2001 have pointed to a number of critical features of patriotism that makes it a controversial issue when discussing citizenship, and in particular the relation between the individual and the country's values (e.g.Mummendey, Klink and Brown, 2001;Condor, 2001;Hopkins, 2001).Such controversies include the relation between patriotism and nationalism (e.g.Kosterman and Feshbach,1989;Blank, Schmidt, and Westle, 2001), the "good" or "bad" nature of these concepts, (e.g.Janowitz, 1983;Connor, 1993), the conditions under which patriotism and/or nationalism induce out-group derogation (Blank et al., 2001;Mummendey, Klink and Brown, 2001), and the stable and fundamental nature of patriotism.The present article does not address all of these problems, but rather focuses on the latter: the stability of patriotism.This is, in our view, a problem for citizenship.More specifically, the stability of patriotism is critical when dealing with the ability of the democratic alternation (the alternation in power of parties that hold different values and goals, typical of democratic countries) to induce different representations of the country, as well as different forms of attachment to it.Indeed, if patriotism is a stable feature for the members of a country, for example if it is acquired through the exposure to national history and iconography (Reykowski, 1997), then it should show little sensitivity to variations in the values expressed by the different groups, parties and coalitions that alternate in power in a democracy.If this is true-and we will argue against it-it may be a problem, because it would imply that attachment to and representation of the country depend only upon long-term historical factors and not upon the political activity of the State, such as national decisions, the orientation of a specific government (that in most democratic countries is in power "only" for four-five years), and the discourses of politicians.

The stability of patriotism
The recent literature on the roots of patriotism seems to argue for a stable "nature" of patriotism.For instance, Johnson (1997) argues for an evolutionary approach, considering patriotism as an inborn predisposition that applies to all social systems and is insensitive to short-term incentives.In this view, patriotism is considered as a basic and fundamental force in "human nature", anchored in such evolutionary mechanisms as nepotism and reciprocity (Buss, 1987).In the same vein, Feshbach and Sakano (1997) argue that attachment to the country is a fundamental process, although they find its origin in the generalization of affect from parents to nation: because the nation fulfills for the adult the same needs as the ones that parents fulfill for the child, attachment to the country is understood as a deep loyalty to, literally, the "fatherland".Indeed, these authors found empirical support for this assumption by showing that patriotism is correlated with attachment to the father.It is interesting to note that this position is quite similar in structure to the one defended by Adorno, Frenkel-Brunswik, Levinson and Sanford in 1950, when authoritarianism was considered a stable personality trait, rooted in the parents' education.In spite of the abundant criticism that this approach has generated (see for example Brown, 1996), the temptation to consider social and political life as definitively determined by early interaction with parents is still widespread.
Historical approaches to patriotism (e.g.Ben-Amos, 1997) also tend to consider the development of patriotism as based on non-contingent factors.In this view, patriotism is determined by such factors as the teaching of national history in schools, or the building of monuments for past heroes; it follows that patriotism is bound to remain a quite rigid construct that can only evolve over several generations.
In sum, there is a significant tendency in the literature to conceive patriotism as a stable construct.This might lead to a deterministic conception of citizenship, in which citizens of a certain country are supposed to react in a pre-determined way, as far as factors affecting attachment to the country (e.g.aggression, relations to immigrants) are concerned.
The present article proposes that such a view conceals the fact that people's relations to the country, even affective relations such as attachment, are political relations, and therefore should be affected by the political activity of the State.This was proposed by Branson (2002): "The citizen who truly loves this country must first understand and be freely committed to its ideals-the values and principles to which it is committed.The citizen who truly loves this Republic must also be willing to take part in safeguarding those values and principles and in helping to effect the more complete realization of this nation's ideals.In short, the citizen must be both 'enlightened' and 'engaged'".In other words, patriotism should be partly related to the political state of affairs.
This view is consistent with Habermas' concept of "constitutional patriotism" (1996), a form of attachment to a constitutional order, i.e. a particular national interpretation of more general constitutional principles, such as human rights.The idea that patriotism can be related to high-order political considerations is also pointed out by the notion of "enlightened political socialization" (Habermas, 1996), which relates to the fact that culture, and cultural byproducts such as patriotism, are not reproduced dogmatically through blind attachment to tradition (Hendley, 1998), but rather through active political relationships.Moreover, Habermas' ideas are in line with Sandel (1982) and Taylor (1989) who propose that the national sense of community stems from a common appreciation of the support for liberal (in the classic sense) political principles.Thus, even if patriotism can be considered as an attachment to the nation based on an affective dimension, it also appears important to consider the influence of political factors.This, in a way, is in line with Habermas' (1992) distinction between the ethnic-cultural and the democratic sides of citizenship, also called Ethnos and Demos dimensions of national citizenship (see also Sanchez-Mazas, van Humskerken, and Casini, 2003).In sum, the above work concurs with the argument in favor of the variability of patriotism, at least on the basis of relevant political factors.
Based on this reasoning, the first, general, hypothesis is that patriotism should be influenced by the political decisions taken by the State.In the research that will be presented, French nationals were exposed to experimentally different laws proposed by the parliament, as a proxy for the variations in political decisions of the State.
The structure of patriotism If patriotism is not merely a stable personality trait, but the result of political life, then several kinds of patriotism might be conceived, depending on how individuals envision the social and political systems in which they live.Accordingly, Schatz (1995;Staub, 1997;Schatz & Staub, 1997;Schatz, Staub, & Lavine, 1999) distinguished between blind and constructive patriotism.The main difference between these two forms lies in the nature of attachment.Whereas blind patriotism refers to a rigid and inflexible attachment to the country, characterised by a loyalty without any criticism-and this regardless of the nature of the group's behaviour-, constructive patriotism is more flexible.It refers to a "critical loyalty" of current group practices, driven by a desire for positive change (see also Fletcher's "enlightened loyalty", 1993).Whereas blind patriotism has no consideration for the values that do not belong to the group, constructive patriotism considers humanitarian value as of fundamental worth.
A study carried out by Schatz (1995, in Schatz & Staub, 1997) assessed blind patriotism through a three-item scale strongly suggesting that the country should not be criticized ("I would support my country right or wrong"; "People should not constantly try to change the way things are in America"; and "It is un-American to criticize this country"), and constructive patriotism through a three-item scale asking for critical involvement ("My love of the country demands that I speak out against popular but potentially destructive policies"; "People should work hard to move this country in the right direction"; and " If you love America you should notice its problems and work to correct them").Analyses showed that whereas constructive patriotism was positively correlated with empathy and pro-social values, the reverse was observed for blind patriotism.Also, blind patriots obtained a higher score on the symbolic obedience dimension than did constructive patriots.Furthermore, constructive patriots expressed a higher score on the implication and political activism dimensions.In another study, Schatz et al. (1999) showed-here with nineteen items-that blind, but not constructive, patriotism was positively associated with the belief that symbolic behaviours were more important to the country than instrumental ones.In sum, these authors pointed out that blind and constructive patriotism represent two qualitatively distinct forms of positive identification with the country and affective national attachment.This distinction is useful for addressing the problem of stability: Although we proposed that political variations should influence patriotism, it is also true that some elements of patriotism are linked to more long-term factors such as the history and culture of one's own country.Thus, it is possible to think that blind patriotism is a more rigid construct, because it refers to symbolic attachment (national anthem, statues of heroes, and so on; cf. the correlation with symbolic obedience), and that constructive patriotism is a more flexible construct, because it refers to political involvement (improvement of society, evaluation of government's action, and so on; cf. the correlation with political activism).The latter is also implied by Staub (1997), when presenting constructive patriotism: "Since policies, practices, institutions, and cultures are always imperfect, and require change as social conditions and circumstances in the world change, and since national policies and practices are conducted by governments made up by individuals who have varying worldviews […], the capacity and willingness by citizens to engage in corrective actions is essential" (p.214).

Hypotheses
The second, more specific, hypothesis is that, if it is true that patriotism is a flexible construct (as stated in the general hypothesis), this should be true only for its constructive form.It is worth noting at this point that "blind patriotism" and "constructive patriotism" are probably not the best terms for these constructs.From a strictly descriptive, operational, point Stability of Patriotism 9 of view, they should be called "uncritical" and "critical" attitudes toward the country's activities.When taking into account their correlates, they could be called respectively "symbolic attachment" and "political involvement".For now, we will keep Schatz and Staub's terms for the sake of comparison; alternatives will be addressed in the conclusion.
The present study assessed the expression of blind and constructive patriotism after the exposure to the political orientation, in terms of internal policy decisions, of the national parliament.The script exposed participants to a law, recently voted by the parliament.This law concerned the allocation of an extra budget, and was either egalitarian (concerned nationals and immigrants), pro-ingroup only (discriminatory), or pro-outgroup only (threatening).The blind and constructive tendencies are therefore used in this study as dependent variables.This is a paradigmatic change compared to previous studies on patriotism that looked at it as a predictor, but this is also-and more importantly-a theoretical difference.Using patriotism as a dependent variable derives from the hypothesis that patriotism can vary, and this is the first attempt to provide empirical evidence for this hypothesis.
Whereas blind patriotism refers to a rigid loyalty to the country and a sensitivity to symbolic attachment, constructive patriotism is supposed to be more flexible, as it is linked to an ability to criticize the country.Accordingly, the operational hypothesis was that blind patriotism should be stable regardless of the values underlined by the law, whereas constructive patriotism should vary.

Participants
The total sample consisted of 61 French citizens, 41% males and 59% females.Their age ranged from 19 to 80 (M = 41.92,SD = 19.41).The participants were approached by the experimenter in public places (cafeterias, parks, etc.) where they were asked if they had a few Stability of Patriotism 10 minutes to answer an anonymous questionnaire presented as a "Study on Relationships between Groups".All of them were volunteers.Among them, 19.7% were students, 52.5% were employed, 18% were retired and 9.8% were unemployed.

Materials and procedure
First, participants had to write down their nationality and also the country to which they felt the most attached: Individuals who answered a country other than France were excluded from the analyses.They completed some demographic items.
Second, participants were exposed to a fictitious law that had been allegedly voted by the French parliament a few days earlier.It was mentioned that the media had been filled with a lot of spectacular news in the past days, and thus this information had been overlooked by many citizens.Participants were asked to read the following text carefully: "The law 1412B concerning the rights of the needy citizens […] has just passed the approval of the Senate House, after the approval of the Chamber of Deputies.This law consists of an official recognition of the needy citizens […].This law provides supplementary help for the integration of the needy persons […] through a supplementary tax, corresponding to 0.1% of the global income taxes.This law should be applied starting from January 2001."The parliament was chosen as the influence source, in order to use a source that might be considered as representative.
Three kinds of laws were proposed by manipulating the recipients of the supplementary help (inserted in the above text, in the location of the "[…]").(a) An egalitarian law: The tax will benefit needy persons, "French nationals as well as immigrants".
(b) A pro-ingroup-only law: The tax will benefit needy persons, "with priority of French nationals".(c) A pro-outgroup-only law: The tax will benefit needy persons, "with priority of regular immigrants".This between-subjects manipulation was intended to represent the variability of attitudes that the legislative institution (here, the parliament) can hold towards matters of internal politics such as, in this case, the (institutional) attitude towards immigrants.
Participants were asked to rate the law in terms of agreement, morality and legitimacy, on 13-point scales ranging from 1 (negative pole) to 13 (positive pole).They also expressed to what extent they viewed the parliament as representing the national authority, on 13-point scales, ranging from 1 (not at all) to 13 (totally).
Finally, participants filled in the two scales that constituted the main dependent variables: (1) A 6-item blind-patriotism scale, assessing the participants' orientation toward blind patriotism.Participants expressed their opinion on a 6-point Likert scale (1 = totally disagree, 6 = totally agree) regarding these statements: "I will support my country right or wrong"; "It is un-French to criticize this country"; "When I see the national flag, I do not feel concerned"; "It is important that the national French unity has priority over the European unity"; "It is important to maintain the national ceremonies, such as the 14 th of July"; "Individuals who do not stand up when listening to the national anthem are disrespectful".This scale yielded a good internal consistency (Cronbach's alpha = .74)in the present study.
(2) A 5-item constructive-patriotism scale, assessing the participants'orientation toward constructive patriotism.Participants expressed their opinion on a 6-point Likert scale (1 = totally disagree, 6 = totally agree) regarding these statements: "The fact that I am a French citizen is an important part of my identity"; "People should work hard to move the country in a good direction"; "If one loves France, one should notice its problems and work to correct them"; " I do like my country, but before that, the most important thing is the well-being of all individuals, whoever they are"; "My love of the country demands that I speak out against all potential governmental policies that could be demagogic".This scale yielded an acceptable internal consistency (Cronbach's alpha = .64)in the present study.The above items are quite similar to those used by Schatz (1995) in order to allow at least a theoretical comparison, but have been adapted to the French context.
It is important to note that these two scales have been validated in a previous study (Depuiset & Butera, 2002, study 2).This study showed good internal consistency for both scales (Cronbach's alpha = .87for blind patriotism, and Cronbach's alpha = .73for constructive patriotism), and allowed the researchers to point out a number of correlates.It appeared that blind patriotism correlated positively with dogmatism (Rokeach, 1956), with the expression of negative stereotypes towards immigrants, with feelings of relative economic deprivation, and with identification with the town and the region (and also with France, but this was a ceiling effect); it also correlated negatively with authoritarianism-rebellion (Kohn, 1972), with positive attitudes towards immigrants, and with identification with Europe.
Conversely, constructive patriotism did not correlate with dogmatism; it correlated negatively with the expression of negative stereotypes towards immigrants, with feelings of relative economic deprivation, and with identification with the town and the region; it also correlated positively with authoritarianism-rebellion, with positive attitudes towards immigrants, and with identification with Europe (again, positive identification with France was a ceiling effect).
After completing the questionnaire, participants were thanked and thoroughly debriefed.

Agreement with the laws
First of all, it should be noted that overall participants thought that the parliament represents the national authority (M = 7.68).This is true regardless of the experimental conditions; the ANOVA on the representativeness item did not yield any significant effect, F (2, 56) = .65,p = .53.Thus, the influence source seems to be considered a legitimate one.

Stability of Patriotism 13
As for agreement, the mean scores obtained on agreement, legitimacy and morality were aggregated in a single agreement score (Cronbach's alpha = .95).An ANOVA on this score showed an effect of the type of law, F(2, 58) = 5.14, p = .009.As shown in Table 1, participants agreed more with an egalitarian law (M = 11.17)than with a pro-outgroup-only law (M = 7.22), t(58) = 3.21, p = .002.Agreement with the pro-ingroup-only law (M = 9.17) appeared to be in between, but was not significantly different from the other two.
- -------------------------------Insert Table 1 around here --------------------------------Expression of blind and constructive patriotism An ANOVA on the blind patriotism scale showed no effect of the type of law, F(2, 58) = .35,p = .70;overall mean = 3.67.As discussed in the introductory section, blind patriotism refers to symbolism and attachment to the country, which could explain its stability and its independence from contingent prescriptions.On the contrary, constructive patriotism seemed to vary according to the type of law, F(2, 58) = 3.08, p = .054.As shown in Table 2, participants expressed more constructive patriotism when exposed to a pro-ingroup-only law (M = 5.00) than when exposed to an egalitarian one (M = 4.34), Tukey HSD post hoc test, p = .042.In other words, considering the nature of the constructive items, participants agreed more with the importance of expressing some criticism and some political involvement under a discriminatory law, rather than under an egalitarian one.The other differences were not significant.

Discussion
The present study proposes some advances in the research conducted on patriotism, and more particularly on the distinction between blind and constructive patriotism.First, it contributes to the external validity of the blind and constructive patriotism scales, as it did not use a student sample like previous studies did.Second, it was conducted in a country other than the United States, which may be relevant for the study of these concepts themselves.These two points had been suggested by Schatz et al. (1999) as important future issues for research on patriotism.However, contrary to Schatz et al. (1999), we did not replicate the negative correlation between the two constructs.In our study, they were either independent or positively correlated.It is important to note that in the above mentioned study, Depuiset and Butera (2002) found a negative correlation (r = -.60), but it was much stronger than that found by Schatz and collaborators; in other studies, not yet formalized in a manuscript, they found sometimes negative and sometimes positive correlations.It thus appears that the two constructs are not necessarily exclusive, and that their relation changes when the experimental materials change.This could suggest the importance of considering various contexts when dealing with patriotism.It is interesting to note that the same idea is pointed out, at a macrosociological level, by Blank et al. (2001): "the observed historical link in the earlier genesis of nation-states between ethnic-national identity and democracy is to be interpreted as a contingent parallelity, but not as a systematic necessary connection".
More importantly, this study revealed that blind and constructive patriotism were not equally stable: Whereas blind patriotism-referring to symbolic and affective attachment towards the nation-seemed to be stable and resisted the influence of the context, constructive patriotism varied in our study.Because the latter refers to personal and political implications, this may explain why it can be expressed differently depending on specific political contexts, especially in the face of laws that express important choices with regard to social and political issues (welfare and acceptance of immigrants).These results are in line with Tamir's (1997) suggestion that patriotism may vary depending on the changes in the perception of interests, norms and social structure.Interestingly, participants expressed a higher degree of constructive patriotism when exposed to a pro-ingroup-only law.This is only speculative, but if the expression of constructive patriotism is the sign of personal and political involvement, then the latter result can be interpreted as a kind of "vicarious guilt", a sort of compensation due to the fact that the reference group has displayed a discriminatory, anti-normative behavior.Some authors have discussed the possibility of feelings of guilt at the group level, as a consequence of the behaviour of other in-group members (e.g.Baumeister & Hastings, 1997;Skevington, 1989).Doosje, Branscombe, Spears and Manstead (1998) found that members of a group can experience guilt because of the past behavior of their in-group toward another group, even when they personally played no role in harming the out-group.
Furthermore, these feelings of guilt apparently motivated compensation for the past behavior of their group.Further studies including assessments on collective guilt, as well as the relation between patriotism and guilt are needed.
Finally, it is interesting to note that the agreement with the egalitarian law is very high, as a sort of truism; it appeared to be significantly higher than agreement with the prooutgroup law, which might seem evident, since the latter could be seen as threatening for the ingroup.Agreement with the pro-ingroup-only law, which is in fact a discriminatory norm, lies in between: Not quite different from egalitarianism, but not quite different from ingroup threat either.It might be an overgeneralization, but it looks like an illustration of the ambivalence of attitudes toward discrimination.

Conclusions
The present study clearly does not solve all the problems raised in the introductory section.Patriotism is a complex phenomenon and it needs to be studied with multiple perspectives.However, the above results allow us to make two comments.
The first one concerns the stability of patriotism.The present study suggested that patriotism is not a stable construct with a fixed structure.Whereas blind patriotism seems to be resistant to variations in the political context, constructive patriotism is reactive to it.This, in our view, represents an important contribution for the understanding of patriotism.The above data showed that the stability of patriotism only concerns symbolic attachment to the country, attachment to timeless icons of nationality such as the flag, the national anthem and the national celebrations.Indeed these symbols are clearly detached from the current and practical contingencies of political debate, and it seems reasonable that they are more resistant to change.However, it is possible to argue that our experimental manipulation questioned a political problem, and that it is understandable that it would be unlikely to affect a symbolic construct.An interesting question that might be asked, and tested, is whether it would be possible to produce a variation in blind patriotism.Indeed, there are times in which symbols are directly questioned by the political activity; these are generally times of revolution, in which statues, portraits and monuments are broken down or burned, but they can also be times of profound regeneration, like when a state works at the writing of a new Constitution (e.g.Italy's recent passage from the First to the Second Republic).Such crises might be able to elicit variation in the symbolic attachment typical of blind patriotism; future research will determine the plausibility of such a hypothesis.
Conversely, our experimental manipulation affected constructive patriotism.Social influence is a matter of relevance (Butera & Mugny, 2001), and the manipulation of national decisions such as new laws is highly relevant for influencing a construct based on political involvement.Such a view is supported by social influence studies showing that change is linked to the correspondence between the type of knowledge or attitude under examination and the type of message used by the source in its communication (e.g.Buchs, Falomir, Mugny, & Quiamzade, 2002;Butera, Mugny, & Buchs, 2001;Pérez & Mugny, 1996).Thus, together with the above argument on the possibility of affecting blind patriotism, we are in fact suggesting that variations in patriotism are linked to the relevance of social and political communications to which citizens are exposed: Communications on the organization of social life (such as institutional attitudes towards immigrants) will affect constructive patriotism; communications on the foundations of the country (such as a new constitution) could affect blind patriotism.In such a way, patriotism can be understood as the result of political communication, and not as an inborn mechanism.Again, further research is needed to anchor this suggestion with empirical results.
The first comment leads to the second.When writing about patriotism in this article, we have frequently referred to blind patriotism as "symbolic attachment" and to constructive patriotism as "political involvement".These terms derive from the content of the items used, by Schatz and colleagues, and by us, to compute the two constructs, as well as from their correlates, but they also stem from our understanding of patriotism as determined by social and political communication.We do not know at this stage if they have to be substituted for the terms "blind patriotism" and "constructive patriotism", but they should at least make us reflect upon what we call patriotism."Symbolic attachment" and "political involvement" refer to different levels of social and political involvement, the former being linked to fundamental values of the country, the latter related to more operational attitudes; only in this sense, and not considering it as a stable construct, can patriotism be understood as an active mechanism of citizenship (see also Staerklé, Roux, Delay, Giannettoni, & Perrin's, 2003, differentiation between consensus-based and conflict-based citizenship).

Table 1 :
Agreement with the law.Means that do not share the same subscript are significantly different at p < .05.Agreement ranges from 1 (not at all) to 13 (totally).

Table 2 :
Expression of constructive patriotism.Means that do not share the same subscript are significantly different at p < .05.Constructive patriotism ranges from 1 (not at all) to 6 (totally)