Sanctions as a Tactic Used in Partner Conflicts

Abstract

Partner sanction in this study is a form/tactic of violence, much like verbal and physical violence, which partners use toward each other during their conflicts. The partner sanction embodies a temporary deprivation of a mutually agreed-on right. The purpose of this study is to develop a theoretical and operational framework of sanctions partners use. The study sampled 74 heterosexual couples from the general population (148 male and female participants). The findings support the validity and reliability of the sanction measurement. Furthermore, findings indicate that the use of sanctions between partners is highly prevalent among men and women in the general population; that the more one partner uses sanctions, the more the other partner uses it; and that sanctions are strongly associated with other violent tactics partners use in their conflict (i.e., verbal and physical). Theoretical and empirical implications of the theoretical framework and the findings are discussed, including the role of sanctions in partner conflicts that escalate to severe forms of violence.

Keywords

partner violence sanctions verbal aggression psychological abuse physical attack conflict escalation gender symmetry

This study addresses the sanctions intimate partners use toward each other. A sanction used by one partner toward the other is defined in this study as a temporary deprivation of a mutually agreed-on right, wherein the partner using the sanction is aware that depriving the other partner from this right may hurt him or her, and is using it as part of an attempt to cope with a conflict. Quite surprisingly, sanction use by intimate partners toward each other received no significant or in-depth theoretical or empirical attention. Questions such as what is a sanction between intimate partners, how it is expressed, what it is used for, what makes it unique, what it has in common with other actions intimate partners use toward each other in their conflicts, and how often it is used, were not discussed or examined nor answered in previous relevant literature. The purpose of the present study is, first, to present a theoretical framework of sanctions to enable an answer to these questions; then, to suggest an operational approach to measuring sanctions; and finally, to examine it and provide preliminary reliability and validity indices.

The study of partner sanction use may contribute to understanding and coping with various aspects of PV. The following examples are non-exhaustive and will be further dwelled on: First, the study of sanctions extends the focus of interest from physical expressions of PV to the broader context which includes non-physical violence as well. Second, the study of sanctions alongside more severe forms of violence makes it possible to examine not only its behavioral aspects but also its dynamics on the levels of relationship and conflict, such as what are the characteristics of violent partner relationships, and how conflicts in such relationships evolve and escalate to physical violence with high hurt potential. Third, the study of sanctions allows for a reevaluation of basic concepts that have become static in the field. These examples are sufficient to maintain that the study of partner sanction use could have implications for the perceptions, policy, services, and intervention methods that are acceptable in the field.

Sanctions in the Context of Violence

The first question to be presented and examined to characterize partner sanctions is whether there is a foundation to regard and classify sanctions as a form of PV? To answer this question an agreed definition of violence is needed to enable such assertion. The World Health Organization (WHO; 2002) defines violence (not necessarily physical or PV) as follows:

The intentional use of physical force or power, threatened or actual, against oneself, another person, or against a group or community, that either results in or has a high likelihood of resulting in injury, death, psychological harm, maldevelopment or deprivation. (p. 5)

Like many other definitions, this one posits two prerequisites for an action to be classified as violence: that the action used by the actor has the potential to harm, and that he or she has an intention to harm others (Winstok, 2006). Relatively speaking, it is simpler and more straightforward to determine whether there is harm potential than whether there is harm intention. Clearly, sanctions are potentially harmful, thus meeting the first prerequisite, but what about the second prerequisite? It is hard to tell. It is possible that sometimes one partner deprives another of a privilege not intending to harm but rather finding it impossible to bestow that privilege in the situation between them (conflict). For example, even if it is agreed that the partners usually express love toward each other, during a fight that is very difficult to actualize. When one or both partners are hurt, angry and frustrated, it would be very difficult for them to exhibit sympathy, affection, and closeness. Not expressing intimacy by one partner to the other is not necessarily avoidance aimed to hurt but may also be temporary distance brought about by conflict. The difficulty to determine intent did not stop numerous researchers from assigning hurtful intentions to hurtful actions, classifying them as violence. For example, yelling and even cursing were classified as verbal violence (Straus, 1979, 1990). Can it be unequivocally ascertained that when one partner raises his/her voice and even yells or curses, it is done with an intention to hurt the other partner? Not necessarily. It is possible to yell at things, curse the weather, self, and others not to hurt them but as a spontaneous means of venting and relieving frustration. In this context, it is worth noting that one of the early theories of violence, designated Aggression–Frustration theory, addressed the positive relationship between these concepts (Dollard, Doob, Miller, Mowrer, & Sears, 1939). This difficulty to determine intent is sometimes used by individuals who hurt their partners. To reject their classification as violent and to minimize consequent societal response they claim lack of intention to hurt (Eisikovits, Goldblatt, & Winstok, 1999). If hurtful actions of unclear and difficult to classify intention are defined as violence, as with yelling and cursing, then sanctions should also be defined as violence. In light of the limitations presented earlier the above definition of sanctions deliberately avoids addressing intent, referring only to the sanctioning partner’s knowing that the sanctioned partner may get hurt. Therefore, the approach taken here is that if one party in the intimate conflict knowingly hurts the other party, intentionally or unintentionally, it is violence.

This argument renders knowing that an action is hurtful a central notion, and reduces hurtful intent to a marginal idea, in the context of intimate relationships. This approach significantly diminishes the ability of individuals who knowingly behave hurtfully to manipulate their responsibility for their actions. For example, a man who deprives his wife from accessing their joint bank account knows that this action hurts her. He may maintain that this forceful action bears no hurtful intent, but rather is an attempt to protect family interests. According to the intent criterion, such action by the man is not considered to be a violent one. However, according to the knowing criterion, the man’s action is considered violent. In the present study, knowing overtakes intent.

The next question to be examined is where to place sanctions in the context of common classifications of violent actions (e.g., verbal and physical violence) taken between intimate partners? For that purpose, the common classifications of PV in the literature should be examined. Violence forms between partners are classified in numerous, not always coherent, ways. One of the widespread classifications in the field was used by Straus and colleagues (Straus, 1979, 1990; Straus, Hamby, Boney-McCoy, & Sugarman, 1996) in developing the most prevalent measurement of PV, Conflict Tactic Scales (CTS). The theoretical framework Straus and colleagues used to develop the CTS stemmed from conflict theories (Adams, 1965; Coser, 1956; Dahrendorf, 1959; Scanzoni, 1972; Simmel, 1955; Straus, 1979). According to these theories, violence is a non-legitimate tactic individuals use to settle interpersonal conflicts in general, as well as within the family and between intimate partners. The CTS was adopted by many researchers in the field of family violence and became a widespread standard instrument for measuring PV. The first version (CTS-1) was introduced in 1979 (Straus, 1979, 1990), the second (CTS-2) in 1996 (Straus et al., 1996). Straus and his colleagues as well as other researchers used this instrument mainly to measure men’s violence against their female partners and vice versa, triggering a debate over whether violence on the part of the two sexes can be addressed in the same manner (Winstok, 2012). The CTS consists of five scales representing various tactics the authors considered relevant to the study of partner conflict. Only two of the five scales, physical and non-physical, are discussed because they received much attention and are widely used by researchers.

Straus and colleagues (Straus, Gelles, & Steinmetz, 1982) define physical violence (CTS-1 used the term physical aggression; CTS-2 used the term physical attack) as an act carried out with the intention of or perceived intention of causing physical pain or injury to another person. The physical attack scale differentiates between mild and severe attacks. Mild actions include object throwing, arm twisting, hair pulling, pushing, holding or confining movement, and slapping. Severe actions consist of using a knife or gun, punching, kicking, strangling, throwing against a wall or floor, beating, and burning. The non-physical violent behaviors scale was designated verbal aggression in CTS-1 and as psychological aggression in CTS-2. The Psychological aggression scale in CTS-2 also differentiates between mild and severe behaviors. Mild actions include yelling, cursing, and avoidance; severe actions were insults, threats, and vandalism. Straus (1979) referred to psychological abuse as verbal and non-verbal acts which symbolically hurt the other, or the use of threats to hurt the other. Obviously the definition of psychological abuse is quite vague as compared with the definition of physical violence (synonymous with physical aggression or physical attack), which is no coincidence.

Most violence definitions focus on the action rather than its outcomes (Winstok, 2006). Examples are the general WHO definition presented earlier and the following specific definition by Straus and colleagues of physical violence, which also vaguely define non-physical (verbal and/or psychological) PV. Whereas the CTS-1 terms physical aggression scale, CTS-1 verbal aggression scale, and CTS-2 physical attack scale that replaced the CTS-1 term physical aggression are pretty consistent with widespread violence definitions, the CTS-2 term psychological aggression scale that replaced the CTS-1 term verbal aggression scale is not. The notion “psychological” does not refer to the action but rather to its outcomes (consequences or effects)—despite the fact that it measures behaviors rather than outcomes. Simply put, the term “psychological aggression” is inappropriate in representing violent behaviors. Why then did Straus and colleagues exchange an appropriate term (verbal aggression scale in CTS-1) for an inappropriate one (psychological aggression scale in CTS-2)? The authors explained the change citing some actions the scale measures as non-verbal. It seems that the authors of the CTS were faced with three options: limit the scale to verbal violence only (omitting non-verbal actions); define the scale as non-physical violence. In the first case, the scale would have been focused on and limited to verbal violence, and in the second case, the scale would have been defined by something it is not rather than by the thing it is. The authors chose a third option—to define the scale imprecisely. The scale is mainly verbal and includes some non-verbal and non-physical actions. Despite the limitations of the choice the authors made, the loose definition is not that crucial. The title Straus and colleagues gave the measurement, and the theoretical framework they founded it on, indicate that they were more interested in producing scales that measure conflict tactics used by intimate partners than in scales that measure PV. The fact that their measurement includes negotiations supports this idea (Straus et al., 1996). The sanction definition presented earlier is consistent with the CTS authors’ approach as it also considers sanctions as a tactic used by partners in their intimate conflicts (i.e., using it as part of an attempt to cope with a conflict), rather than as a form of violent behavior.

The decision made by Straus and colleagues to present non-physical behaviors as psychological aggression is not unique. Numerous scholars adopted the term psychological violence as a means to include various non-physical actions in PV (regardless of whether they are violent actions or conflict tactics; see O’Leary, 1999). Although psychological violence had also been called psychological aggression, psychological abuse, emotional aggression, emotional violence, and emotional abuse (McHugh, Rakowski, & Swiderski, 2013), the meaning is still the same. As already argued, the term psychological aggression is a major deviation from the basic definition of violence because it is not focused on the action but on its consequences. As such, it allows adding to PV undefined behavioral categories with negative psychological consequences without having to prove hurtful intention. Straus and colleagues resorted to using this term to be able to include in their scale, alongside the verbal violence, additional non-verbal violent behaviors such as “stomped out of the room,” without needing additional theoretical classifications. Despite their cautious and limited use of the term psychological violence, the term is a slippery slope that may serve to bypass the prerequisites of the definition of violence.

It is widely agreed and acknowledged that the term psychological aggression lacks an accepted definition and accordingly is quite blurred and vague (Follingstad, 2007; O’Leary, 1999). O’Leary (1999) argued that psychological abuse is represented by acts of recurring criticism and/or verbal aggression toward a partner, and/or acts of isolation and domination of a partner. Generally, such actions cause the partner to be fearful of the other or lead the partner to have very low self-esteem, and it is recommended that researchers in this area routinely assess the impact of psychological abuse. Marshall (1996) provided a complementary argument in the same direction, suggesting that psychological abuse targets the victim not through physical abuse, but through non-physical means including derogation, ridicule, coercion, humiliation, disrespect, intimidation, threats, vandalism, surveillance, and control. O’Leary and Marshal present different lists for the same classification, once more exposing the term’s theoretical and practical shortcomings.

Types of Sanctions in the Context of Partner Violence

This study is based on a distinction between two types of sanctions. One type is when an intimate partner temporarily deprives another partner to some degree of a right to satisfy emotional needs such as appreciation, affection, and love in their relationship. This sort of sanction is typically expressed in emotional coldness, distance, and minimal verbal and physical interaction on the part of the sanctioning partner toward the sanctioned one. This type of sanction is used mostly spontaneously; it will be referred to as “spontaneous sanction” from this point on. Another type of sanctions is when one partner temporarily deprives another partner to some degree of a right to satisfy physical needs, for example avoids performing important customary acts such food preparation or purchase of certain items. This type of sanction is of a punitive nature; it will be referred to as “punitive sanction” from this point on. The difference between the two sanction types is not only in their orientation but also in their focus and scope of possible maltreatment. Both types of sanctions have an emotional impact, but whereas in the first type this is the major focus, in the second it is but one of two emphases. The other emphasis is on denying a desire that is habitually fulfilled. The first type of sanction is simple, unstructured, and is spontaneously used, therefore it is reasonable to expect that it is more common than the second, punitive and more structured type. This statement is based on the general principle that simple behaviors tend to be used more often than complex ones. This argument was not empirically tested in the field of PV, yet evidence for it can be found in the methods used by parents in educating their children. Apparently, the chronicity and prevalence of deprivation of privileges are lower than those of verbal violence (Fauchier & Straus, 2007), likely because the latter is relatively more simple, unstructured, and spontaneous than the former. It is reasonable to place both types of sanctions in close proximity with verbal violence, classified also as “psychological violence.” One significant difference between the sanctions and the other forms of PV is that while other violence types such as verbal or physical violence refer mainly to actions that should not be performed, sanctions mainly refer to actions that should be performed but are not. Another noteworthy difference is that compared with other violent actions, the sanction can be presented as an irreproachable, non-violent act.

The basic definition of sanctions, their structure, and classification in the context of PV as presented earlier is insufficient to complete the theoretical framework toward their operationalization and examination. It is still necessary to determine how this form of violence is associated with other forms of PV. Numerous scholars address the various violence types between partners as tactics used to cope with conflicts (Winstok, 2012). Since sanctions are also a type of violence exhibited in partner conflicts, they are a tactic as well (refer to the initial definition of sanctions in the beginning of this article, which regards sanctions as a means of coping with conflicts).

Numerous studies found an association between the uses partners make of the various violent tactics. The most conspicuous examples are found in studies that identify a relationship between psychological and physical violence (Hines & Saudino, 2003; Krebs, Breiding, Browne, & Warner, 2011; Sullivan, McPartland, Armeli, Jaquier, & Tennen, 2012; Winstok & Perkis, 2009). O’Leary (1999) mentioned additional examples of this relationship to establish the importance of psychological violence as a factor predicting physical violence. Hence, sanctions can also be expected to be closely linked to other forms of PV.

Empirical evidence of the linkage between violent tactics used by partners in conflict is important but insufficient to characterize the relationship of sanction types to other tactics. An alternative theoretical framework to violence is required. The CTS hints at such alternative framework: partner conflict.

Sanctions in the Context of Conflict Escalation to Violence and Escalation of Violence in Conflicts

Partner conflicts may arise when one partner experiences a discrepancy between his or her view of desirability and availability, and perceives the other partner as responsible for this gap. However, not every such gap would necessarily create a conflict. At times, such gaps might facilitate a reconsideration of expectations, which in turn, narrows the gap so that no conflict is generated. A conflict might evolve when a discrepancy appears to require intervention and correction. When conflicts do arise, they can be resolved by normative means and can also remain unresolved. Not all conflicts must be settled. Numerous couples have normative, satisfying relationships even without complete agreement. Yet, not all conflicts are conducted or straightened out peacefully. Sometimes, one or both partners try to force their position on the other through non-legitimate means (i.e., violence; Winstok, 2012).

The above-mentioned experienced discrepancy could serve as the onset of a possible conflict that might lead to PV. Focus on this process could explain how partner conflicts arise, evolve, and end—with or without violence. This approach places emphasis on the situational context of actions taken in conflict, including violent behaviors. It does not maintain that violence results solely from a situation, but that understanding the situation is highly relevant for understanding the violence. Although the significance of the process through which violence arises has been recognized, and the theoretical and empirical basis of the process is established in other fields dealing with interpersonal violence, it has received little attention in the study of PV (Winstok & Perkis, 2009).

PV is typically the outcome of an escalatory conflict: At the outset, the conflicting parties tend to use normative (non-violent) actions to resolve the conflict (e.g., explanation, negotiation, and compromise). But when these actions fail to resolve the conflict and at least one party insists on a resolution, actions may become increasingly hurtful and belligerent. The purpose of such behavioral escalatory dynamics is to sustain relative conformity (Winstok, 2006, 2008, 2012; Winstok & Eisikovits, 2008).

Conformity, in the context of PV, is obtained through minimal, if any, deviation from the common social norms in two aspects: the goals of the conflicting parties, and the means they use to achieve those goals. Goals and means possess value. As a rule, goals are conceptualized in terms of benefit or gain, and means in terms of cost or loss. More often than not, a maximum cost–benefit ratio is sought for. This can be obtained if the goals are achieved using the least costly means. The process of escalation enables that and also explains the relationship between and frequency of use of partner tactics in conflict. At the outset of a conflict, the parties use tactics with the lowest personal, interpersonal, and social cost—such as mild aggression—trying to achieve their goals. As low cost tactics fail, higher-cost tactics are then used, such as severe aggression—as long as the value of the goals is higher than the cost of the means to achieve them. Based on these principles, the escalatory dynamic is produced, manifested in the relationship between the tactics and the frequency the partners use them toward each other in the conflict.

Two criteria can help in positioning the violence forms on cost continuum (escalation): the action’s effortlessness and convenience, and the action’s negative consequences for the aggressor and the victim (i.e., risk). The more a violent action is effortless and convenient and the less it has negative consequences, it would tend to position itself toward the beginning of the escalatory continuum. As mentioned by O’Leary (1999), psychological violence, including verbal aggression, precedes physical violence on the escalatory continuum (the former predicts the latter). Both are relatively effortless and convenient, but the former poses less negative consequences than the latter and therefore precedes it. What about spontaneous and punitive sanctions though? The spontaneous sanction is rather effortless and convenient, and is probably less risky than the punitive sanction. Therefore, it is reasonable to assume that the former precedes the latter. The emerging escalatory continuum begins with the spontaneous sanction (due to its considerable effortlessness, immediacy, and low-risk factor) and ends with physical violence (due to its high-risk factor). Between the two, it is reasonable to assume that punitive sanction will follow the spontaneous one and precede the verbal violence, since it is less risky. The positioning of the violence forms on the escalatory continuum provides the basis for their relationships: The correlation between adjacent violent forms on the continuum is higher than between distant forms. The positioning of the violence forms on the escalatory continuum also relates their incidence: The first forms are more prevalent than the last. This would mean that the spontaneous sanction is the most common, punitive sanction and verbal violence are less so, and the least common is physical violence.

Thus far, the discussion focused on the question how violent behaviors of one partner escalate (one behavior affects the next one of the same individual), without accounting for the effect of the other partners’ behaviors on his or her behavioral tendencies. That is an important question albeit a non-exhaustive one, as it focuses on the individual aspect of escalatory dynamics. There is however another, more important aspect, which is the interactive one. In that case, the focus is on the question how violent behavior by one partner affects violent behavior by the other partner in a relationship, and vice versa. The distinction between the individual and interactive aspects is used for heuristic purposes aiming to simplify a complex process. In reality, both aspects motivate the development of partner conflict and at times produce escalatory dynamics. In most cases, conflict is interactive rather than individual. Escalatory dynamics is influenced by the actions both partners take. Each new action by any partner in the relationship is affected by at least two information inputs: what the research participant did (e.g., cursed his or her partner) and how the participant’s partner responded (e.g., the partner cursed the participant back). Therefore, not only the link between a partner’s actions should be examined but also the link between each partner’s actions to those of the other one. The actions of the other partner can decrease or increase the escalation process.

How is sanction use integrated in the interactive process? Does sanction use moderate or augment the escalatory process, and how? As suggested earlier, the tactics are set on an escalatory continuum. Sanction tactics are positioned at the start followed by verbal and then physical violence tactics. It is known that the tactics used by both partners in conflict are positively correlated and that not all conflicts escalate to violence (Winstok, 2012). Therefore, it seems that a spontaneous sanction used by one partner will increase the probability that the other partner will also use it (interactive aspect). A punitive sanction used by one partner would increase the other partner’s use of it (interactive aspect). A spontaneous sanction used by one partner could potentially increase that same partner’s use of a punitive sanction (individual aspect). It remains to be postulated how one partner’s use of spontaneous sanction would affect the other partner’s use of punitive sanction (interactive aspect). Sanctions are positioned at the beginning of the escalatory continuum and their activation may mark the onset of an escalatory process. Therefore, when one partner uses sanction, it may deter the other partner from stepping into an escalatory conflict, effectively terminating the conflict. Thus, one partner’s use of spontaneous sanction moderates the other partner’s use of punitive sanction (interactive aspect). Figure 1 presents the associations between sanctions used by intimate partners proposed here.

Figure 1.

Association between partners’ uses of sanctions.

Gender and Partner Violence

The relationship between gender and PV received much attention and raised heated debate in the field (Winstok, 2011, 2012, 2013). In light of the importance of gender in this context, the question of how it affects sanction use should be considered as well. Yet lack of research on sanctions necessitates finding an answer to this question via the effect of gender on other types of PV. A large body of empirical evidence unequivocally supports the notion that a significant (but not necessarily equal) proportion of both genders use violence in their intimate relationships (see review, Langhinrichsen-Rohling, Selwyn, & Rohling, 2012). This notion was designated “Gender Symmetry” (Straus, 2011; Winstok, 2011, 2012). Although the accumulated empirical knowledge consistently supports the idea of gender symmetry in perpetration of PV, it is important to emphasize that most studies show that women suffer more injuries than men (Jose & O’leary, 2008). The obvious conclusion from studies of gender symmetry in physical PV is that it exists also in cases when the partners use more moderate types of violence toward each other, including the sanctions addressed by the present article.

Despite the significant importance assigned to differences between men and women in PV, the comparison depends on the actions it refers to. The actions characteristic of each of the partner conflict tactics can differ. For example, physical violence includes nail scratches and hair pulling; as it would be impossible to pull on the hair of a bald person or scratch with very short fingernails, such actions are more likely to be taken by those who have longer nails or against those with longer hair. Therefore, it would seem that in partner conflicts, nail-scratching is a violent actions more characteristic of women whereas hair pulling more of men. If physical violence were represented and measured by one-gender-typical actions, the measurement would have shown that this gender was using physical violence more frequently than the other. To avoid such biased measurement, the gender basis of the variety of violent actions (i.e., sample of behaviors) by which each tactic including sanctions between heterosexual couples should be considered in the operationalization of each tactic, or else artificial asymmetry may be produced. But taking in consideration the gender basis of actions representing tactics may hinder an examination of gender differences in tactic use, producing artificial symmetry. Symmetry and asymmetry may be biased artifacts produced by biased operationalization.

Summary of Research and Hypotheses

At this point, sanctions in this study were characterized in a manner allowing the development of a preliminary measurement of sanctions, as well as the drawing of hypotheses to assist in testing the validity and reliability of this measurement. The measurement is based on an operationalization of the definition and the theoretical characterization of sanctions as presented thus far with the same approach that characterized the development of the physical and verbal aggression CTS scales. The sanction measurement proposed here can be regarded as an additional distinct scale of the CTS. The definition and characterization of the sanctions serve as basis for the following main hypotheses, examined in the present study:

Hypothesis 1: An examination of the factor structure of sanction measurement will reveal two distinct types of tactic: spontaneous and punitive.

Hypothesis 2: According to the position of the tactics on the escalatory continuum stemming from their effortlessness, convenience and risk, it will be found that:

The frequency of spontaneous sanctions use and their user rate will be highest, followed at decreasing frequency and rate by verbal violence, punitive sanctions and physical violence.

The association between frequencies of use for the adjacent tactics on the continuum will be higher than that between non-adjacent tactics (individual aspect of escalation).

Hypothesis 3: Gender symmetry will be found in the frequency and rate of tactic use:

Frequency and rate of use by men and women of each tactic will be similar.

An association will be found between frequency and rate of partners’ use of tactics: The more one partner uses a tactic, the more the other partner would tend to use that tactic (interactive aspect of escalation).

Method

Data Collection Principles and Sample Description

As this study aims to provide initial support to a new measure, it is based on a snowball, non-representative convenience sample which is common in this type of studies (similarly to the sample used in testing the validity and reliability of the revised CTS; Straus et al., 1996).

Cohabiting heterosexual couples were recruited to this study using the snowball method, and were consecutively separately interviewed. At the beginning of a data collection session, the surveyors presented orally and in writing the purpose of the study and its importance, the issues the participant will be asked to address and how he or she is asked to report them. Prior to beginning the data collection procedure, the participant rights were presented both orally and in writing, including anonymity, confidentiality, the right to refuse to participate, to stop at any point and not to answer part of the questions. In this setting also the surveyors’ obligations were presented, including confidentiality, participant anonymity, and use of the data for research purposes only. Then the participant was handed a questionnaire to fill out. The participant was informed that he or she is welcome to approach the surveyor for clarifications and assistance. When both partners in the relationship completed the questionnaire, it was placed with no identifying marks in an envelope.

The sample consisted of 74 heterosexual couples from the general population (148 participants: 74 males and 74 females). Men were in the age range of 30 and 50 years, 41.01 years in average (SD = 6.7), and the women were in the age range of 28 and 50 years, 38.43 years in average (SD = 5.51). As for education, 12.2% of men and 4.1% of women had partial high-school education, 27.0% of men and 37.8% of women had completed their high-school education, and 60.8% of men and 59.81% of women had higher academic (i.e., BA and MA) education. Of the sample, 96% of men and women reported working part-time or more, and 4% reported not working; 8.1% of men and 6.8% of women reported above average economic status; 83.8% of men and 82.4% of women reported average economic status; and 8.1% of men and 10.8% of women reported a below average economic status.

Measurement Instrument

Four measurements were used in this study: two of partner sanctions (two new subscales) and two of partner aggression (two CTS-2 scales).

Partner Sanction Subscales (New Measurement)

Based on the theoretical framework presented earlier for sanctions and the CTS measurement method, a measurement of spontaneous and punitive sanction subscale was developed. The difference between the two is in the type of actions research participants were asked to refer to. Both scales were presented in tandem to research participants. The introductory comments to questions stated the following:

Sometimes, women and men eschew giving or doing for their partner something that they usually give or do. Such avoidance is often in response to a behavior by the partner that they find unacceptable, offensive or hurtful. The following questions refer to such cases.

Each subscale consists of 12 items. Six items address the sanction taken by the respondent’s partner toward him or her (victimization items) and 6 items address the same sanction taken by respondent toward his or her partner (perpetration items). Research participants were presented with a series of sanctions, for each of which they were asked to evaluate “How often over the past year did your partner behave as detailed in this column in response to your unacceptable, offensive or hurtful behavior? And how often did you behave that way toward your partner?” The five response options for each question are as follows: (a) never happened in the past year or at all; (b) never happened in the past year, but did happen prior to that; (c) happened once or a few times over the past year (low frequency); (d) happened once or a few times per month (moderate frequency); (e) happened once or a few times per week (high frequency). For each subscale (i.e., spontaneous and punitive sanctions), two types of new variables for men’s use of sanctions and women’s use of sanctions separately were encoded (2 subscales × 2 genders × 2 types of variables resulted in 8 new variables). The new variables were based on men and women’s combined reporting on selves and partner (each partner reported on self and partner’s use of sanctions). One type of variable differentiates between two categories of PV (dichotomous): no violence (not using sanctions) and violence (using sanctions). This type of variable was used for prevalence analysis. The second type was based on all scale items’ average (interval). This type of variable was used for chronicity analysis. This scoring procedure is consistent with the recommendations of the CTS authors (Straus et al., 1996).

Spontaneous sanction subscale

The sanctions addressed in this subscale are as follows: behaved toward partner in one or more of the following manners: indifferently, uncaringly, sulking, evaded talking to them; refrained from supporting partner or approving of them even when they deserved it; abstained from expressing physical affection to partner, such as a kiss, a caress, or a hug; rejected partner’s attempts to be nice to you; refrained from calling partner by their pet name; did not talk to partner, ignored them, and did not respond to their questions or requests. Measurement reliability in for this subscale in this study is highly acceptable (internal consistency using α Cronbach was .87 for men and .88 for women).

Punitive sanction subscale

The sanctions addressed in this subscale are as follows: did not make something that you usually do for partner, such as a cup of coffee, a cake, dinner; did not perform something that you usually do for partner, such as driving them where they need to go; prevented partner from buying stuff they need, want or like, such as a food or a clothing item; avoided, were late to, or objected to accompanying partner to an event important for them such as a family, social, or work-related event; did not buy partner stuff they need, want or like, such as a food or a clothing item; have gone missing on partner for several hours or did not respond to their phone calls. Measurement reliability in this study is highly acceptable (internal consistency using α Cronbach was .84 for men and .83 for women).

Conflict Tactic Scales (CTS-2)

The study used two CTS-2 scales developed by Straus et al. (1996). The two scales were psychological aggression and physical attack. The difference between the two is in the type of actions the research participants were asked to refer to. Both scales were presented in tandem to research participants. The introductory comments to questions stated the following:

Regardless of how well partners get along with each other, it happens that disagreement arises, partners get angry with each other, want something from one another, or a fight somehow transpires. The following list describes behaviors that can happen when there is disagreement.

The psychological aggression scale consists of 8 items and the physical attack subscale consists of 22 items. In each scale, half of the items address the actions taken by the respondent’s partner toward him or her (victimization items) and half of the items address the same actions taken by respondent toward his or her partner (perpetration items). Research participants were presented with a series of actions, for each of which they were asked to evaluate “At what frequency over the past year did your partner and you behaved toward each other in the following manner.” The following five response options to each question are: (a) never happened in the past year or at all; (b) never happened in the past year, but did happen prior to that; (c) happened once or a few times over the past year (low frequency); (d) happened once or a few times per month (moderate frequency); (e) happened once or a few times per week (high frequency). The scoring procedure for the CTS scales was identical to the one presented for the sanctions subscales, and it was consistent with the recommendations of the CTS authors (Straus et al., 1996). For each scale, two types of new variables were encoded for men and women separately (2 scales × 2 genders × 2 types of variables resulted in 8 new variables). The new variables were based on men and women’s combined reporting on selves and partners (each partner reported on self and partner’s use of sanctions). One type of variable differentiates between two categories of PV (dichotomous): no violence (using no sanctions) and violence (using violent tactics). This type of variable was used for prevalence analysis. The second type was based on all scale items’ average (interval). This type of variable was used for chronicity analysis.

Psychological Aggression Scale (as Verbal Aggression Scale)

The actions addressed in this scale were as follows: insulted or swore at partner, called partner fat or ugly, shouted at partner, threatened to hit or throw something at partner. The following four items that are on the original scale were excluded from this study for improperly representing verbal violence: stomped out of room, destroyed something of partners, did something to spite partner, accused partner of being a lousy lover. Measurement reliability in this study is acceptable (internal consistency using α Cronbach was .70 for men and .83 for women).

Physical attack scale

The actions addressed in this scale were as follows: threw something at partner that could hurt; twisted partner’s arm or pulled hair; pushed or shoved partner; hit partner with something; choked partner; slammed partner against wall; beat up partner; grabbed partner; slapped partner; burned or scalded partner on purpose; kicked, bit, or punched partner. The “used knife or gun on partner” action that appeared on the original scale was excluded from this study because of the small probability this behavior would have been found. The introductory comment, response options, and encoding of the new variables were identical to those used for the psychology aggression scale. Measurement reliability in this study for women was acceptable (internal consistency using α Cronbach = .89). For men, due to the low levels of physical violence, and as for some items all the sample reported “never happened” with zero variance (the items were not excluded from the scale), measurement reliability as internal consistency was relatively low (α Cronbach = . 60). But this result is not unique (i.e., Rivera-Rivera, Allen-Leigh, Rodriguez-Ortega, Chavez-Ayala, & Lazcano-Ponce, 2007; Straus & Mickey, 2012).

Data Analysis Principles

The first stage, after computing and recoding research variables, is to test the first hypothesis and examine the factorial structure of the sanctions measurements for men and women separately. The second stage is to test the second hypothesis and examine the prevalence and chronicity of all conflict tactic forms in a comparable manner. This is done by analyzing the co-occurrence of men’s use of conflict tactic forms toward women, women’s use of it toward men and men and women’s use of it toward each other; analyzing the effects of gender and conflict tactics forms on the chronicity of their use; and analyzing the correlations between men’s conflict tactic forms, women’s conflict tactic forms, and men and women’s conflict tactic forms toward each other. The third stage of data analysis aims to deepen the understanding of the relationships between the conflict tactic forms by exploring the individual and interactive escalatory patterns using path analysis. In all the analysis stages, the third hypothesis referring to gender symmetry issues is also examined.

Results

Factorial Structure of the Sanctions Measurements

To test the factorial structure of the measurement, principal component analysis was conducted using a Varimax Rotation method with Kaiser Normalization for two-factors solution. Choosing a two-factor solution was primarily based on the theoretical framework which stated that sanctions can be sorted to spontaneous and punitive. Further on, the scree plot also supported a two-factor solution. Table 1 presents the factor loadings of the men and women’s test separately. The test reinforces the theoretical distinction (two sanction forms) for men and women. The test indicates that the six items representing spontaneous sanctions load on one factor for both genders and the other six items representing punitive sanctions load on the second factor for both genders. Loadings were very good (Factor Loadings > .59). This test supports the first and the third research hypotheses.

Table 1.

Two Factor Solution Loadings for Men and Women.

	Factor Loadings
	Men^a		Women^b
Item Description	Spontaneous Sanction	Punitive Sanction	Spontaneous Sanction	Punitive Sanction
You behaved toward your partner in one or more of the following manners: indifferently, uncaringly, sulking, evaded talking to them	.836	.256	.773	.233
You refrained from supporting your partner or approving of them even when they deserved it	.806	.118	.732	.076
You abstained from expressing physical affection to your partner, such as a kiss, a caress, or a hug	.763	.295	.874	.196
You rejected your partner’s attempts to be nice to you	.706	.319	.742	.417
You refrained from calling your partner by their pet name	.671	.170	.664	.213
You did not talk to your partner, ignored them, and did not respond to their questions or requests	.655	.294	.712	.305
You did not make something that you usually do for your partner, such as a cup of coffee, a cake, dinner	.281	.841	.411	.650
You did not perform something that you usually do for your partner, such as driving them where they need to go	.023	.820	−.006	.756
You prevented your partner from buying stuff they need, want or like, such as a food or a clothing item	.241	.807	.154	.835
You avoided, were late to, or objected to accompanying your partner to an important event for them such as a family, social or work-related event	.354	.606	.336	.593
You did not buy your partner stuff they need, want or like, such as a food or a clothing item	.384	.558	.332	.696
You have gone missing on your partner for several hours or did not respond to their phone calls	.290	.531	.285	.633

Note. Consistent with scree plot result. Extraction method: principal component analysis; rotation method: Varimax with Kaiser Normalization.

Eigenvalues = 1.497.

Eigenvalues = 1.519.

Prevalence of Conflict Tactic Forms

Figure 2 presents the prevalence of men and women’s conflict tactic forms. The prevalence of the spontaneous sanctions is the highest for both men and women, while it is higher for women (94.6%) than for men (93.2%). The prevalence of punitive tactics and verbal aggression is lower than that of spontaneous tactic. For men, the prevalence of both tactics (verbal aggression = 89.2%; punitive = 82.4) is slightly higher than that for women (verbal aggression = 86.5%; punitive sanction = 86.5%). The prevalence of physical violence is the lowest for both men and women. The prevalence of this tactic for men (33.8%) is higher than that for women (32.4%). However, the difference in prevalence between genders is small and not significant (95% confidence interval). The differences between spontaneous sanction, punitive sanction, and verbal aggression are small as well and not significant (95% confidence interval). The large and significant difference is between the latter three and physical aggression.

Figure 2.

Prevalence of men and women’s conflict tactics (percentage).

Tables 2 to 4 present the co-occurrence of the various tactics. Table 2 presents the co-occurrence of men’s use of conflict tactic forms toward women, Table 3 presents the co-occurrence of women’s use of conflict tactic forms toward men, and Table 4 presents the co-occurrence men and women’s use of conflict tactic forms toward each other.

Table 2.

Co-Occurrence of Men’s Use of Conflict Tactics Toward Women (Percentage).

	Verbal Aggression		Punitive Sanctions		Physical Attack
	No	Yes	No	Yes	No	Yes
Spontaneous sanctions
No	2.7	4.1	5.4	1.4	6.8	0.0
Yes	8.1	85.1	12.2	81.1	59.5	33.8
Verbal aggression
No			5.4	5.4	10.8	0.0
Yes			12.2	77.0	55.4	33.8
Punitive sanctions
No					17.6	0.0
Yes					48.6	33.8

Table 3.

Co-Occurrence of Women’s Use of Conflict Tactics Toward Men (Percentage).

	Verbal Aggression		Punitive Sanctions		Physical Attack
	No	Yes	No	Yes	No	Yes
Spontaneous sanctions
No	2.7	2.7	4.1	1.4	5.4	0.0
Yes	10.8	83.8	9.5	85.1	62.2	32.4
Verbal aggression
No			5.4	8.1	13.5	0.0
Yes			8.1	78.4	54.1	32.4
Punitive sanctions
No					12.2	1.4
Yes					55.4	31.1

Table 4.

Co-Occurrence of Men and Women’s Use of Conflict Tactics toward Each Other (Percentage).

	Women
	No	Yes
Spontaneous sanctions
Men
No	5.4	1.4
Yes	0.0	93.2
Verbal aggression
Men
No	10.8	0.0
Yes	2.7	86.5
Punitive sanctions
Men
No	10.8	6.8
Yes	2.7	79.7
Physical attack
Men
No	60.8	5.4
Yes	6.8	27.0

The findings indicate that for men and women, the co-occurrence is highest between two adjacent tactics on the escalation continuum, and that the larger the distance between tactics on the continuum, the lesser the co-occurrence. For example, in Table 1 the co-occurrence in men of spontaneous sanctions with verbal aggression (85.1%) and punitive sanctions (81.1%) is higher than with physical violence (38%). Findings also demonstrate that the dyadic co-occurrence of tactics (Table 3) is dominant. This co-occurrence is especially high for spontaneous sanctions (93.2%), verbal aggression (86.5%), and punitive sanctions (79.7%), which are the most commonly used. These findings support the suggestions of escalatory continuum and gender symmetry of various tactics. The findings are inconsistent with the hypothesis that punitive sanctions precede verbal aggression on the escalatory continuum. The findings show that both tactics are positioned on the same area in the continuum. This test supports the second and third research hypotheses. However, alongside testing the question with tactic prevalence, it should also be tested with tactic chronicity.

Chronicity of Conflict Tactic Forms

To examine the effects of gender and conflict tactics forms on the chronicity of these tactics, a multivariate test was used (repeated measures). Table 5 and Figure 3 present the inference and descriptive statistics. These findings are consistent with the prevalence analysis findings.

Table 5.

Effects of Gender and Conflict Tactic Forms on the Chronicity of Conflict Tactics Use (Multivariate Test).

Effect (Independent Variables)	F	df	η²
Gender	0.454	1,73	.006
Form of violence	95.266***	3,71	.801
Gender × Form of violence	5.965***	3,71	.201

p < .05. **p < .01. ***p < .001.

Figure 3.

Chronicity means of men and women’s conflict tactics.

Correlations Between Conflict Tactic Forms

Table 6 presents the correlations between men’s conflict tactic forms toward women, and Table 7 presents the correlations between women’s conflict tactic forms toward men. The findings for men and women demonstrate that the highest correlation is between two adjacent tactics on the escalatory continuum, and that the greater the distance between tactics on the continuum, the smaller the correlation. For example, in Table 6, the spontaneous sanction is correlated with the punitive sanction (.75) more than with any other tactic. It is less correlated with verbal aggression (.52) and the least with physical violence (.34). These findings support the hypotheses of the escalatory continuum and the gender symmetry of the various tactics. Unlike the vague finding in the previous prevalence and chronicity analyses, the findings here are consistent with the hypothesis that the punitive tactic precedes verbal aggression on the escalatory continuum. Table 8 presents the correlations between men and women’s conflict tactic forms toward each other. These findings reaffirm the dyadic correlation in tactic use.

Table 6.

Correlations Between Men’s Conflict Tactics.

	Verbal Aggression	Punitive Sanctions	Physical Attack
Spontaneous sanctions	.52***	.75***	.34***
Verbal aggression		.35***	.48***
Punitive sanctions			.35***

p < .05. **p < .01. ***p < .001.

Table 7.

Correlations Between Women’s Conflict Tactics.

	Verbal Aggression	Punitive Sanctions	Physical Attack
Spontaneous sanctions	.69***	.75***	.21
Verbal aggression		.49***	.62***
Punitive sanctions			.07

p < .05. **p < .01. ***p < .001.

Table 8.

Correlations Between Men’s and Women’s Conflict Tactics Toward Each Other.

	Men–Women
Spontaneous sanctions	.74***
Verbal aggression	.73***
Punitive sanctions	.52***
Physical attack	.23*

p < .05. **p < .01. ***p < .001.

Escalatory Patterns

At this point, the research hypotheses were tested and supported. But the research data allow further in-depth examination of the role of sanctions in an escalatory partner conflict. For this end, path analyses of individual and interactive escalatory patterns were conducted. It is important to note that these analyses could prove problematic due to the strong correlations between the test variables. Such correlations may produce multicollinearity, generating statistical artifacts. Accordingly, the findings of this section of the analyses should be evaluated and interpreted while acknowledging their limitations, and bearing in mind the theoretical framework of the escalatory dynamics of partner conflicts. This issue will be further addressed in the discussion.

Individual and Interactive Escalatory Patterns

So far, the research hypotheses have been supported. However, it is possible to delve deeper into the data to enhance this support and better understand the escalatory pattern using path analysis. In the first phase, the individual escalatory patterns were examined for men and women separately. The analysis results are presented in Figure 4. One model (A) was tested for men, and three models (A, B, C) were tested for women.

Figure 4.

Individual escalatory patterns.

The men’s model is simple and straightforward. Spontaneous sanctions positively affect (enhance escalation) punitive sanctions and verbal aggression. It affects punitive sanctions more than verbal aggression. With the effects of spontaneous sanctions on punitive sanctions and verbal aggression, no correlations were found between punitive sanctions and verbal aggression. Punitive sanctions and verbal aggression positively affect physical violence (escalatory). Verbal aggression has a stronger effect than punitive sanctions on physical violence. With the effects of punitive sanctions and verbal aggression on physical violence, spontaneous sanctions have no effect on physical violence. The effect of spontaneous sanctions on physical violence is mediated by punitive sanctions and verbal aggression.

The model for women is more complex. Women’s Model A is very similar to men’s Model A, but the major difference is that in the women’s model punitive sanctions do not significantly affect physical violence. In Model B, the path between punitive sanctions and physical violence was constrained to “0.” The model produced a negative effect (de-escalation) between spontaneous sanctions and physical violence. The fit indices of this model were the best (see Table 9). Women’s use of spontaneous sanctions not only increases the probability of physical violence use (mediated by punitive sanctions and verbal aggression) but also increases the probability of physical violence non-use. Model C was designed to test whether the negative effect revealed in Model B is not a statistical artifact (multicollinearity). The effect of spontaneous sanctions was completely neutralized in Model C (all the emanating paths were constrained to “0”). The negative effect persisted, this time with an effect of punitive sanctions on physical violence.

Table 9.

Individual Escalatory Patterns’ Model Fit (Sample Size = 74 Couples).

Individual Model	χ²	df	p	NFI	IFI	CFI	RMSEA
Men A	0.0	0	—	—	—	—	—
Women A	0.0	0	—	—	—	—	—
Women B	1.7	1	.20	.99	.99	.99	.09
Women C	89.2	3	.000	.42	.43	.41	.63

Note. NFI = normative fit index; IFI = incremental fit index; CFI = comparative fit index; RMSEA = root mean square error of approximation.

Interactive Escalatory Patterns

The second phase, following an examination of the individual path, tested the interactive path. Preliminarily, it was tested only for the two sanction types and then, as described next, it was tested for all the tactics addressed in this study. The interactive effects between partner sanctions were tested by means of three models presented in Figure 5 and Table 10: Model A is the central one while Models B, C, and D are used to assist in understanding Model A results. The results of Model A testing indicate that from the interactive aspect, each partner’s use of spontaneous sanctions is correlated with the other partner’s use of spontaneous sanctions: The more one partner uses them, the more the other partner does, too. From the individual aspect, the use each partner makes of spontaneous sanctions positively affects that partner’s use of punitive sanctions (escalation enhancing effect). The use of spontaneous sanctions increases the probability that punitive sanctions will be used by each partner separately. Back to the interactive aspect, each partner’s use of the spontaneous sanctions tactic has a negative effect (de-escalation effect) on the other partner’s use of punitive sanctions tactic. The use one partner makes of spontaneous sanctions increases the probability of punitive sanctions non-use by the other partner. To ensure that the negative effects are not a statistical artifact (multicollinearity) Models B, C, and D were tested. The correlation between men and women’s spontaneous sanctions was constrained to “0” in Model B, the effect of men’s spontaneous sanctions was neutralized in Model C (all its associations were constrained to “0”), and in Model D, the effect of women’s spontaneous sanctions was neutralized (all its associations were constrained to “0”). Constraining the correlation between men’s and women’s spontaneous sanctions to “0” in Model B did not change the negative effects found in Model A as expected. Furthermore, the neutralization of spontaneous sanctions for men in Model C and for women in Model D turn the negative effects into positive as expected. Hence, it seems that one partner’s use of spontaneous sanctions directly decreases the other partner’s use of punitive sanctions yet indirectly increases it.

Figure 5.

Interactive escalatory sanctions patterns.

Table 10.

Interactive Escalatory Patterns Model Fit (Sample Size = 74 Couples).

Individual and Interactive Model	χ²	df	p	NFI	IFI	CFI	RMSEA
A	0.0	0	—	—	—	—	—
B	57.6	1	.000	.75	.75	.75	.88
C	147.3	3	.000	.37	.37	.36	.81
D	148.8	3	.000	.36	.36	.36	.82

Note. NFI = normative fit index; IFI = incremental fit index; CFI = comparative fit index; RMSEA = root mean square error of approximation.

The Interactive Escalation Model

Figure 6 presents the model that includes all the tactics tested in this study. The model presents the associations (effects and correlations) between tactics partners use during their conflicts. Such tactics include: spontaneous sanctions, punitive sanctions, verbal aggression, and physical assault. The model differentiates between individual effects (the behavior of one partner has an effect on another behavior of that partner) and interactive effects and correlations (the behavior of one partner has an effect on or correlation with the behavior of the other partner). It is important to differentiate between escalation enhancing (positive) effects and de-escalatory (negative) effects.

Figure 6.

Interactive escalatory patterns.

The Effects of Using Spontaneous Sanctions

The findings regarding the effects of using spontaneous sanctions are identical with those presented above. Interactively, each partner’s spontaneous sanctions use is correlated with the other’s use of spontaneous sanctions: The more one partner uses them, the more the other does, too. Individually, each partner’s use of spontaneous sanctions tactic positively affects that partner’s use of punitive sanctions tactic. Using spontaneous sanctions increases the probability that punitive sanctions will be used by each partner separately. Returning to the interactive aspect, each partner’s use of the spontaneous sanctions tactic negatively affects the other partner’s use of the punitive sanctions tactic. One partner’s use of spontaneous sanctions increases the probability of the other partner’s punitive sanctions non-use. Moreover, interactively, each partner’s use of spontaneous sanctions positively affects the other’s use of the verbal aggression tactic. Hence, the use of spontaneous sanctions by one partner increases the probability that verbal violence will be used by the other partner. Individually, among women there is an addition that is absent from among men. Women’s use of spontaneous sanctions also raises their probability of using verbal violence. That is not the case with men. Each partner’s use of spontaneous sanctions is correlated with the other partner’s use of them. Each partner’s use of spontaneous sanctions raises his or her probability of using punitive sanctions (escalatory), raises the probability that the other partner will not use punitive sanctions (de-escalatory), and raises the probability that the other partner will use verbal aggression (escalatory). As mentioned, women’s use of spontaneous sanctions raises their probability of using verbal violence as well (escalatory).

The Effects of Using Punitive Sanctions

Interactively, each partner’s use of punitive sanctions is correlated with the other partner’s use of punitive sanctions: The more one partner uses these sanctions, the more the other partner does as well. Individually, only among women, using punitive sanctions negatively affects their use of physical violence (de-escalatory). Interactively, only among men, using punitive sanctions positively affects women’s use of physical violence (escalatory). Each partner’s use of punitive sanctions is correlated with the other partner’s use of such sanctions as well. Punitive sanctions by both partners affect physical aggression among women only. Women’s use of punitive sanctions raises the probability of their non-use of physical aggression, and men’s use of punitive sanction increases the probability that women will use physical aggression.

The Effects of Verbal Aggression Use

Interactively, each partner’s use of verbal aggression is correlated with the other partner’s use of verbal aggression: The more one uses this tactic, the more the other does, too. Individually, each partner’s use of verbal aggression tactic positively affects that partner’s use of physical violence tactic. Interactively, among men only, the use of verbal aggression negatively affects women’s use of physical violence. Each partner’s use of verbal aggression is correlated with the other’s use of this tactic. Verbal aggression by both partners affects women’s use of physical aggression. Women’s use of verbal aggression increases the probability that they will also use physical aggression (escalatory), and men’s use of verbal aggression raises the probability that women will not use physical aggression (de-escalatory).

Finally, given all the effects and correlations presented above, there is no correlation with each partner’s use of physical violence.

Discussion

The purpose of the study was to present a theoretical framework of partner sanctions, suggest an operational approach for measuring them, and provide initial support for measurement reliability and validity. In conjunction with these major goals, the study also provided the opportunity to pursue additional important goals such as the reexamination of common approaches to PV, beginning with the general definition of PV, through the associations between various forms of violence and their incidence in partner conflicts, to the gender basis of PV. The theoretical research framework, methodology, and results bear theoretical and methodological implications for how researchers address, examine, and understand PV. These implications are discussed next.

Research Implications for the Study of Partner Violence

Of all the above-mentioned goals, the most essential one is that of understanding the relationship between various violence forms and their incidence. But despite its significance, this is but a provisional goal. A more advanced one is to understand the process by which these forms of violence are regulated. The relationship between various violence forms and their incidence probably stems from characteristics of the escalatory process of violent conflicts, which was identified and described in the first part of this article (i.e., sanctions in the context of conflict escalation). It is argued that the escalatory process stems from the motivations of the parties in the conflict. Each party has different goals and means to achieve them. Goals and means possess intrinsic value. There is gain or benefit in achieving goals, while the means to achieve them bear cost. Parties sought to maximize benefits and minimize costs (Campbell, 2005; Haselton & Nettle, 2006; Winstok, 2012). It can be achieved if the least costly means are used to reach the goals. The escalatory process makes that possible and explains the relationship and incidence of tactics used by partners in a conflict. The value of goals and the cost of the means of achieving them are tightly linked to the normative system, which to a great extent dictates them; therefore, the process of escalation can be regarded as an attempt to meet social dictates as much as possible, while deviating as little as possible from the norms.

The forms of violence are the means to identify the existence and function of the escalatory process, but only if these can be positioned on an escalatory continuum guided by two criteria as proposed in the theoretical framework of this article: effortlessness and convenience, and the negative consequences of the action for the aggressor and the victim (i.e., risk). The more a violent action is effortless and convenient and the less it has negative consequences, the more it would tend to position itself toward the beginning of the escalatory continuum. To identify and describe the escalatory process in detail, one or even two means (forms of violence) are insufficient. The greater the number of means, the more detailed the escalatory continuum and accordingly, the description of the escalatory process.

As reviewed in the beginning of the article, the existing body of knowledge on violence proposes only two means that can serve to examine the process of escalation: physical violence, and other, non-physical aggression (Straus, 1979, 1990; Straus et al., 1996). Whereas physical violence is clearly conceptualized, non-physical violence is not. At times, it was presented as verbal aggression, and sometimes broadly and vaguely as psychological violence. One may presume the reason for that was that most researchers of family violence were mostly interested in and focused on physical violence. It seems that some researcher were interested in other violence forms only as a means to advance their understanding of physical violence. An example is O’Leary’s (1999) work mentioned earlier, that recommends promoting the study of psychological violence as a predictor of physical violence. That is also probably one reason why physical violence is a well-drawn, developed, distinct, and clearer concept than the other tactics. Another reason may be that physical violence is the easiest to identify, whereas other forms of violence are relatively less visible. Yet two forms of violence, however clear and distinct, are insufficient for an in-depth scrutiny of the escalatory process. The sanctions addressed in this study provide, among other things, a response to the need for more than two means of examining the escalatory process. As a result of this study, another distinct and researchable form of violence is added to the two commonly studied forms of violence (physical and verbal)—that of sanctions.

The theoretical framework and the findings of this study supports the notion that sanctions are a unique and distinctive group of violent actions and therefore should be considered a highly relevant form of violence to be studied. In addition to being distinctive and unique, this is also a complex form of violence. Findings demonstrate that the sanctions tactic is composed of two specific types: spontaneous and punitive tactics, which are associated with each other and with other violence forms to varying degrees with varying incidence. These are important findings because the sanction tactic effectively contributes two distinct means of describing the escalatory continuum.

Generally speaking the findings seem to support the existence of the escalatory mechanism that links and regulates the different forms of violence. However, in light of these findings, the mechanism is a hypothesized black box, and all one can do is examine its expressions and through them evaluate or speculate on its structure and content. In simple words, this study’s findings provide an entry point into the black box of the mechanism that links and regulates the different violence forms. This is achieved through manifestations of the mechanism as revealed in this and other studies addressing the association between violence forms and differences in their incidence (Hines & Saudino, 2003; Krebs et al., 2011; Sullivan et al., 2012; Winstok & Perkis, 2009).

Deciphering the essence of the mechanism requires more than simply identifying its output. A description of its input is called for as well. It is proposed that the narrower and broader social contexts are to be considered the mechanism’s inputs. This suggestion is not new as it has already been proposed for the study of PV (Wilkinson & Hamerschlag, 2005). Typically, three contexts are identified in PV: the broader socio-cultural context, the narrow personal context, and the interpersonal/situational context between the two partners. The interpersonal context is the one in which the broad and narrow contexts meet and actions are manifested (Winstok, 2006). This is a good starting point in any attempt to understand human behavior, including PV specifically.

The socio-cultural context (the broader context) provides general guidelines as to what is desirable and rejected. But these guidelines are not dichotomous. There are many degrees of desirability and rejection. These guidelines are formed and incorporated into the narrower personal context under numerous constraints such as needs, desires, capabilities, and experiences. For instance, different individuals adopt the socio-cultural dictums of not killing or not stealing in various forms that fit their personality and life realities. Most of humankind would not dispute the duty to uphold these guidelines, yet it can be safely assumed that there will be significant differences in how various individuals perceive and implement them in a given specific society. The same applies to PV. The encounter of personal contexts within the dyad forms the interpersonal dyadic relationship. Conflicts may arise at times, some of which may escalate to violence. It is argued that the mechanism under discussion serves to integrate and adapt (process) the guidelines in various life contexts (input) into partner behavior in dyadic conflict situations (output). The integration and adaptation of these guidelines is the mechanism that forms the relationship between various forms of violence and their incidence.

It was hypothesized in the present study that the relationship between various forms of violence and their incidence are a product of the escalatory mechanism. The findings supported the hypotheses but are insufficient to establish the existence of the mechanism and, specifically, its input. The escalatory mechanism is but one, although not the only, possible explanation for the findings (this issue will be re-discussed in the section on research limitations). Moreover, the attempt made here to expand the findings to characterize the mechanism’s input lacks empirical basis in this study, although it is crucial for further examination of the hypothesized mechanism in future studies (this issue will be re-discussed in the section on recommended future research).

Research Implications for the Study of Gender Differences in Partner Violence

This study follows in the footsteps of numerous studies in the field comparing men and women’s use of various tactics. This examination was conducted bearing in mind the criticism presented in the first part of the article of gender symmetry in PV as a fundamentally false concept. First, it is worth noting that this research provides empirical evidence for gender symmetry in sanction use. This evidence is consistent with the findings of other studies that examined gender differences in other violent tactics in which gender symmetry was found (Langhinrichsen-Rohling et al., 2012). Yet as argued, gender symmetry and asymmetry in PV are false concepts, and measurements can equally be presented to support gender symmetry, asymmetry skewed toward men, and asymmetry skewed toward women. However, this is a sensitive controversial issue that calls for a broader in-depth paradigmatic, theoretical, and empirical discussion (Winstok, 2011, 2012) beyond the scope of this article. In this discussion, the issue is limited to gender differences in the use of sanctions.

The idea that there is symmetry and asymmetry in PV was already raised in the early stages of developing the instrument for sanction measurement. The instrument included additional items which were later removed due to empirical and theoretical incompatibility. The most conspicuous item is one that examined avoidance of sex initiated by one of the partners following a conflict between them. The difference between men and women’s use of this sanction was considerable. First, women used this sanction toward their partners much more frequently than men did. Second, this sanction was part of the spontaneous list for women whereas it was on the punitive list for men. Had this sanction been included in the measurement, it would have produced sanction use asymmetry skewed toward women. This case is an example of how the sample of actions used to represent a tactic bears bias potential for symmetry. Yet it also demonstrates gender differences. It seems that there is limited value in scrutinizing gender differences in violent tactic use and that it may also occasionally prevent learning about significant gender differences.

The criticism of the approach used to study differences between men and women in violent tactic use is not to say that there are no gender differences in PV or that they should not or cannot be studied. There are basic differences between men and women and the sexual sanction is one example. Yet the purpose of studying gender differences in violence use is being criticized. It so happens that more often than not, that type of research aspires to determine who is more violent in intimate relationships: men or women. This question cannot be answered unless a long subjective series of presumptions is adopted, such as the premise that a partner hitting another partner in self-defense is non-violent. It is argued that the question that should be asked is what the gender differences in the use of violence are and what do they stem from, not to promote battles between men and women but to understand genders.

Gender differences in the motivations for using violent tactics are one of the most important focal points in the field. These motivations have been addressed in this discussion as a major input factor in the escalatory process. In light of the importance of the gender-related motivational factor in the escalatory process, it requires a comprehensive theoretical framework. There are various ways to address this issue. The following is one example. Much of the literature on sex differences in aggression is based on two distinctive theories: social role theory (Bettencourt & Kernahan, 1997; Bettencourt & Miller, 1996; Eagly & Steffen, 1986) and sexual selection theory (Archer, 1996; Daly & Wilson, 1988). Sexual selection theory locates the origins of sex differences in human evolution. According to this theory, men along the history of evolution concentrated on reproductive competition, for which status had crucial importance. Women concentrated on pregnancy, birth, and rearing the children, for which security and safety had crucial importance. Social role theory also locates the basis of sex differences in many aspects of behavior in the historical division of labor. Roles produce expectancies about gendered characteristics, leading to different patterns of behavior that are transmitted to future generations through socialization processes. Based on these theories, it is argued that status enhancement is more important for males than for females and is more important for males than risk reduction, whereas the opposite is true for females (Winstok & Straus, 2011). Women’s main concern regarding aggression tends to be risk reduction rather than status enhancement (Campbell, 1999). Status enhancement and risk reduction are two basic gender-related motivations that stem from evolutionary and socialization processes that are the cause and consequence of social expectations, goals, and roles. Notably, the two gender-related motivations we focus on, status enhancement and risk reduction, have a common denominator. Both are evaluated and experienced based on coping with risk. Previous studies of PV (Straus & Winstok, 2013; Winstok & Straus, 2011) examined and supported the argument of different motivations for men and women.

Research Limitations

The present study explored some questions that received little theoretical and empirical attention and its novel nature is a limitation in itself. Yet there are additional limitations that should be addressed.

The sample used in this study is a non-representative convenience one. It consisted of 74 heterosexual couples (148 participants: 74 males and 74 females). It is a small sample and accordingly of limited value in generalizing research findings to the general population. This limitation would have been significant if this study were aimed at presenting the scope of the phenomenon in a given population. However, the purpose of this study was different; it was to present a preliminary measurement of sanctions used between partners and to examine their position among other tactics used by partners in dealing with conflicts. The findings, which supported the hypotheses, provide no more than pilot support to the theoretical and operational framework underlying this study and their relevance to future research.

Although the findings of this study support the validity and reliability of sanction measurement, a single preliminary study of a small indistinct sample is insufficient to determine that the sanctions’ operationalization is valid and its implementation is reliable for different populations. It is worth noting that this measurement, as compared with those of other tactics, is no less and maybe even more sensitive to the socio-cultural context, including gender. This attribute may serve as an advantage as well as a disadvantage. It is an advantage in comparative studies of societies and cultures, and a disadvantage in cross-cultural and cross-societal research. Therefore, using this measurement in different societies may require some adjustments. The measurement in its present form, however, seems to be appropriate for Western societies.

One part of the theoretical framework of this study addresses co-occurrence of violence forms. It is based on a non-causal theoretical framework. A different part of the theoretical framework presents the associations between violence forms as a product of the escalatory dynamic (also referred to as the escalatory mechanism). This part is based on a causal theoretical framework. The hypotheses drawn from the theoretical framework of the study were phrased in a non-causation manner as the research design and data used in this study are cross-sectional. However, the discussion of research findings’ implications on escalatory dynamics is conducted within a causal theoretical framework. It is rightfully argued that drawing causal conclusions based on cross-sectional data without presenting their limitations is undesirable. Hence, it is important to clarify that the discussion of the theoretical implications of research findings on escalatory dynamics should be considered while acknowledging their limitations, being based on cross-sectional data. It is also important to point out the great difficulty in conducting longitudinal studies (precondition for causality inference) focusing on behaviors of both partners in concrete conflicts (see Winstok, 2006). Faced with this difficulty, two options remain: either to not study the subject or to study it within its methodological limitations.

The examined models identified strong associations between independent variables, which may produce statistical artifacts (multicollinearity). Some researchers tend to avoid the presentation of such analyses. In the present study, these analyses could be relinquished without hindering the relevance of the article. The reason why these analyses were reported nonetheless stems from their secondary nature, and because the study successfully addressed the problem of cases suspected of multicollinearity, such concerns are minimized. These analyses seem valid, and with the appropriate cautionary, they may offer more contribution than limitation. It is important to bring this provisional information to the attention of the reader.

Recommendations for Future Research

Generally speaking, the findings of this preliminary study encourage establishing a deeper and broader basis and theoretical framework of sanctions between partners, as well as conducting research to promote understanding of the mechanism linking the various forms of PV. More specifically, what is required first and foremost is to test the sanction measurement instrument in various distinct populations; examine the differences between partners who use sanctions among man-only violent dyads, woman-only violent dyads, and both-partners violent dyads (see description of this typology in Straus, 2014; Straus & Winstok, 2013); identify and examine the socio-economic attributes affecting sanction use; identify and examine the motivations for sanction use between partners; identify and examine the factors affected by sanction use in both partners such as married life quality, number of conflicts, or use of other violence forms. Such studies are likely to advance our knowledge and understanding of PV.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Author Biographies

Zeev Winstok, PhD, Associate Professor at the Faculty of Social Welfare and Health Studies, Center for the Study of Society, University of Haifa.

Ronit Smadar-Dror, Social Worker and a PhD Student in the School of Social Work, University of Haifa.

References

Adams

B. N.

(1965). Coercion and consensus theories: Some unresolved issues. American Journal of Sociology, 71, 714-716.

Archer

(1996). Sex differences in social behavior: Are the social role and evolutionary explanations compatible? American Psychologist, 51, 909-917.

Bettencourt

A. A.

Kernahan

(1997). A meta-analysis of aggression in the presence of violent cues: Effects of gender differences and aversive provocations. Aggressive Behavior, 23, 447–456.

Bettencourt

A. A.

Miller

(1996). Gender differences in aggression as a function of provocation: A meta-analysis. Psychological Bulletin, 119, 422-447.

Campbell

(1999). Staying alive: Evolution, culture, and women’s intrasexual aggression. Behavioral & Brain Sciences, 22, 203-252.

Campbell

(2005). Sex differences in direct aggression: What are the psychological mediators? Aggression and Violent Behavior, 11, 237-264.

Coser

(1956). The functions of social conflict. New York, NY: Free Press.

Dahrendorf

(1959). Class and class conflict in industrial society. London, England: Routlege & Kegan Paul.

Daly

Wilson

(1988). Homicide. New York, NY: Aldine de Gruyter.

10.

Dollard

Doob

C. W.

Miller

N. E.

Mowrer

O. H.

Sears

R. R.

(1939). Frustration and aggression. New Haven, CT: Yale University Press.

11.

Eagly

Steffen

V. J.

(1986). Gender and aggressive behavior: A meta-analytic review of the social psychological literature. Psychological Bulletin, 100, 309-330.

12.

Eisikovits

Goldblatt

Winstok

(1999). Partner accounts of intimate violence: Towards a theoretical model. Families in Society, 80, 606-619.

13.

Fauchier

Straus

M. A.

(2007, July). Dimensions of discipline by fathers and mothers as recalled by university students. Presentation at the International Family Violence and Child Victimization Research Conference, Portsmouth, NH.

14.

Follingstad

D. R.

(2007). Rethinking current approaches to psychological abuse: Conceptual and methodological issues. Aggression and Violent Behavior, 12, 439-458. doi:10.1016/j.avb.2006.07.004

15.

Haselton

M. G.

Nettle

(2006). The paranoid optimist: An integrative evolutionary model of cognitive biases. Personality and Social Psychology Review, 10, 47-66.

16.

Hines

D. A.

Saudino

K. J.

(2003). Gender differences in psychological, physical, and sexual aggression among college students using the Revised Conflict Tactics Scales. Violence and Victims, 18, 197-217.

17.

Jose

O’leary

K. D.

(2008). Prevalence of partner aggression in representative and clinic samples. In O’leary

K. D.

Jose

(Eds.), Psychological and physical aggression in couples: Causes an interventions (pp. 15-35). Washington, DC: American Psychological Association.

18.

Krebs

Breiding

M. J.

Browne

Warner

(2011). The association between different types of intimate partner violence experienced by women. Journal of Family Violence, 26, 487-500. doi:10.1007/s10896-011-9383-3

19.

Langhinrichsen-Rohling

Selwyn

Rohling

M. L.

(2012). Rates of bidirectional versus unidirectional intimate partner violence across samples, sexual orientations, and race/ethnicities: A comprehensive review. Partner Abuse, 3, 199-230.

20.

Marshall

L. D.

(1996). Psychological abuse of women: Six distinct clusters. Journal of Family Violence, 11, 379-409. doi:10.1007/BF02333424

21.

McHugh

M. C.

Rakowski

Swiderski

(2013). Men’s experience of psychological abuse: Conceptualization and measurement issues. Sex Roles, 69, 168-181. doi:10.1007/s11199-013-0274-3

22.

O’Leary

K. D.

(1999). Developmental and affective issues in assessing and treating partner aggression. Clinical Psychology: Science and Practice, 6, 400-414. doi:10.1093/clipsy/6.4.400

23.

Rivera-Rivera

Allen-Leigh

Rodriguez-Ortega

Chavez-Ayala

Lazcano-Ponce

(2007). Prevalence and correlates of adolescent dating violence: Baseline study of a cohort of 7960 male and female Mexican public school students. Preventative Medicine, 44, 477-484. doi:10.1016/j.ypmed.2007.02.020

24.

Scanzoni

(1972). Marital conflict as a positive force. In Scanzoni

(Ed.), Sexual bargaining. (61-102) Englewood Cliffs, NJ: Prentice Hall.

25.

Simmel

(1955). Conflict and the web of group affiliations. Glencoe. IL: Free Press.

26.

Straus

M. A.

(1979). Measuring intra-family conflict and violence: The Conflict Tactics (CT) Scales. Journal of Marriage and the Family, 41, 75-88.

27.

Straus

M. A.

(1990). The Conflict Tactics Scales and its critics: An evaluation and new data on validity and reliability. In Straus

M. A.

Gelles

R. J.

(Eds.), Physical violence in American families: Risk factors and adaptations to violence in 8,145 families (pp. 49-73). New Brunswick, NJ: Transaction Publications.

28.

Straus

M. A.

(2011). Gender symmetry and mutuality in perpetration of clinical-level partner violence: Empirical evidence and implications for prevention and treatment. Aggression and Violent Behavior, 16, 279-288.

29.

Straus

M. A.

(2014). Practical ways to conduct international partner violence research using dyadic analysis. NCFR Report Magazine: Family Focus on International Intimate Partner Violence, Fall, 1-4.

30.

Straus

M. M. A.

Gelles

R. J.

Steinmetz

S. K.

(Eds.). (1982). Behind closed doors: Violence in the American family. Transaction Publishers.

31.

Straus

M. A.

Hamby

S. L.

Boney-McCoy

Sugarman

D. B.

(1996). The Revised Conflict Tactics Scales (CTS2): Development and preliminary psychometric data. Journal of Family Issues, 17, 283-316.

32.

Straus

M. A.

Mickey

E. L.

(2012). Reliability, validity, and prevalence of partner violence measured by the Conflict Tactics Scale in male-dominant nations. Aggression and Violent Behavior, 17, 463-474.

33.

Straus

M. A.

Winstok

(2013, August). Gender differences in the relation of dyadic types of partner violence to depression among university students in 15 Nations. Society for the Study of Social Problems annual meeting, New York, NY.

34.

Sullivan

T. P.

McPartland

T. S.

Armeli

Jaquier

Tennen

(2012). Is it the exception or the rule? Daily co-occurrence of physical, sexual, and psychological partner violence in a 90-day study of substance-using, community women. Psychology of Violence, 2, 154-164. doi:10/1037/a0027106

35.

Wilkinson

D. L.

Hamerschlag

S. J.

(2005). Situational determinants in intimate partner violence. Aggression and Violent Behavior, 10, 333-361.

36.

Winstok

(2006). Conceptualizing and testing an interaction-based unit of analysis for the study of conflict escalation. Journal of Aggression, Maltreatment & Trauma, 13, 37-47.

37.

Winstok

(2008). Conflict escalation to violence and escalation of violent conflicts in intimate relationships. Children and Youth Service Review, 30, 297-310.

38.

Winstok

(2011). The paradigmatic cleavage on gender differences in partner violence perpetration and victimization. Aggression and Violent Behavior, 16, 303-311.

39.

Winstok

(2012). Partner violence: A new paradigm for understanding conflict escalation (The Springer Series on Human Exceptionality). Springer: New York.

40.

Winstok

Eisikovits

(2008). Motives and control in escalatory conflicts in intimate relationships. Children and Youth Service Review, 30, 287-296.

41.

Winstok

Perkis

(2009). Women’s perspective on men’s control and aggression in intimate relationships. American Journal of Orthopsychiatry, 79, 169-180.

42.

Winstok

Straus

(2011). Gender differences in intended escalatory tendencies among marital partners. Journal of Interpersonal Violence, 26, 3599-3617.

43.

World Health Organization. (2002). The World health report: 2002: Reducing the risks, promoting healthy life.