What is Positive, Familiarity or Novelty? The Link between Information Processing and Affect
Sophia Nieuwenhuizen 13 July 2015 Bachelor Thesis Clinical Neuropsychology 6075606 Supervisor: Dr. R.H. Phaf Brain & Cognition
University of Amsterdam
Word Count: 6423 Abstract: 120
2 Contents
Abstract 3
What is Positive, Familiarity or Novelty? 4
Familiarity and Positive Affect 6
Can Novelty Be Positive? 14
A Dual-Route Framework 18
Conclusions and Discussion 20
3 Abstract
Are you attracted to the thrill of the novel, or do you prefer the safety of the familiar? Affective consequences of novel and familiar information processing are subject to debate. The question whether positive affect is elicited by familiarity or by novelty is addressed here. First, a direct link between familiarity and positive affect is explained in the light of an affective monitoring mechanism. Novelty or conflict cause negative affect, whereas a quick resolution of conflict in the form of fluent processing leads to positive affect. Subsequently, it is claimed that novelty can indirectly evoke positive affect by matching to current goals of an activated behavior system. In conclusion, present findings are interpreted in the light of a dual-route framework.
4 What is Positive, Familiarity or Novelty?
Dorothy Custer is no stranger to extreme experiences. Aged 102, this lady BASE jumped off the Idaho Perrine Bridge to celebrate her birthday. The year before, she zip-lined accross the Idaho Canyon. One might say that Dorothy is a novelty seeker: she is always on the lookout for a new thrill.
On the other side of the spectrum we have Karl Pilkington, who favors a familiar and predictable environment. In the television show ‘An Idiot Abroad’, he unwillingly travels around the world. While he is immersed in novel cultures, learns new languages, and tries novel foods, he still jumps on every oppourtunity to emphasize how much better he likes his own home. Whereas Dorothy derives pleasure from novel experiences, Karl does not. How can we explain such a discrepancy?
The idea that we are wired to prefer the familiar over the novel, is widely supported by studies of mere exposure (Zajonc, 1968; for a review, see Bornstein, 1989). The basic premise of the mere exposure effect is that repeated, unreinforced exposure to a stimulus will enhance liking for this stimulus, causing a preference for familiarity over novelty. What distinguishes novelty from familiarity is that we do not have an existing representation in memory of a novel stimulus, whereas we do of a familiar one. This has implications for the ease and speed with which a stimulus can be processed, also referred to as fluency (Winkielman & Cacioppo, 2001). When a novel stimulus is encountered, the perceiver is required to extract various pieces of information about it (i.e. name, shape, meaning) and create a new mental
representation of it, which is relatively effortful and complex. When a familiar stimulus is encountered, however, it merely needs to be matched to the already existing mental
representation in order for all information about the stimulus to become available (Claypool, Hall, Mackie, & Garcia-Marques, 2008). Thus, whereas novelty is characterized by low processing fluency, familiarity is characterize by high processing fluency.
5 Fluency appears to be directly linked to affect, which is either positive or negative, and is not necessarily conscious (Phaf & Rotteveel, 2012). Typically, high fluency is linked to positive affect, whereas low fluency is linked to negative affect. In sum, familiarity entails high fluency and positive affect, whereas novelty entails low fluency and negative affect. Whereas the link between familiarity and positive affect is very robust, the mechanism behind this phenomenon is still subject to debate.
Surprisingly, a link between novelty and positive affect has also been suggested. For example, we seem to prefer novel and complex objects over simple ones (Berlyne, 1970; Bornstein, Kale, & Cornell; 1990), seek novelty to increase arousal (Pliner & Loewen, 2002), and prefer it over familiarity when we are in a happy mood (De Vries, Holland, Chenier, Starr, & Winkielman, 2010).
This lack of consensus on the affective valence of familiarity and novelty processing leads to the following question: Is positive affect elicited by familiarity or by novelty? In order to answer this question, it is essential to understand the link between information processing and affect.
In the first section, I argue that familiarity is characterized by fluent processing and that both familiarity and fluency are directly linked to positive affect. Several theories that try to explain the familiarity-positive link will be discussed. However, familiarity is not positive in all situations. It appears that initial conflict is a prerequisite for the elicitation of affect, which is the basic premise of the affective monitoring hypothesis by Phaf and Rotteveel (2012). Affective monitoring directly illustrates how novelty is negatve, and a quick change to fluency is positive.
In some situations, novelty can over time become preferred over familiarity. In the second section, I will argue that novelty can indirectly elicit positive affect through activation of evolutionarily prepared behavior systems. Individual differences in affective traits, like
6 boredom proneness (Bornstein et al., 1990), arousal (Pliner & Loewen, 2002), and even mindset or motivational state (Hansen & Topolinski, 2011) can act as moderating factors in novelty or familiarity preference, and will be discussed in the light of these behavior systems. In the final section, I propose that the seemingly contradictory effects of memory processing on affect can be explained in the light of a route framework. Such a dual-route framework has formerly been proposed by Oatley and Johnson-Laird (1987), who distinguished between a direct (affective) route and an indirect (cognitive) route of processing. While the direct route is characterized by fast, global processing, the indirect route is characterized by analytical and more detailed processing. Whether novelty or familiarity evokes positive affect could therefore depend on the mode of processing that is being used. A dual-route framework could account for mixed findings in familiarity and novelty preference.
Familiarity and Positive Affect
Familiarity and positive affect are closely linked. From an evolutionary perspective we may be wired to prefer the familiar over the novel (Bornstein, 1989). It is safer to eat familiar foods than something potentially poisonous, and it is safer to walk along a familiar path than to take a new, potentially dangerous route. In this secton, competing theories that might explain this phenomenon will be discussed. I will argue that affective monitoring offers the most basic, yet complete framework for understanding the link between information processing and affect.
The idea that we evaluate familiar objects more positively than novel objects has been widely supported by studies of the mere exposure effect; a term coined by Robert Zajonc in 1968 (see Bornstein, 1989 for a review). However, it was initially unclear whether this
7 preference reflects an affective reaction. Therefore, it could be explained by affective, as well as cognitive theories.
Such cognitive theories (i.e. Mandler, Nakamura & Van Zandt, 1987) argue that fluency is enhanced by prior presentation and then misattributed to stimulus liking. Therefore, they predict that evaluations of a stimulus can be either positive or negative, depending on the context. Affective theories, on the other hand, argue that familiarity directly evokes a genuine affective reaction, independently of further cognitive processing. Therefore, they predict that evaluations of merely-exposed stimuli will be only positive.
Harmon-Jones and Allen (2001) found support for the affective explanation of mere exposure. In their study, 37 participants were supraliminally exposed to photographs of female faces. The first set of stimuli contained photographs of 10 different faces, and each face was exposed five times. In the second set, 10 new faces were added into the mix. Positive affect was measured through self-reported liking for the pictures, and through facial
electromyographical (EMG) measurements of the zygomaticus (“smiling”), and the
corrugator (“frowning”) muscles. Results showed that previously exposed faces were rated as more likable and evoked more zygomaticus region activity than novel faces. While
zygomaticus region activity increased with repeated exposures, corrugator region activity was not infuenced. As EMG data provides sensitive and real-time measures of emotional
responses, it is a reliable indicator of positive and negative affect (Larsen, Norris, & Cacioppo 2003). This indicates positivity of familiarity, rather than negativity of novelty. It was
concluded that familiarity evokes genuine positive affect. As exposures were supraliminal and could thus be consciously perceived, cognitive two-step theories cannot be ruled out
completely. However, latencies associated with EMG better match quick affective processing than the generally slower mode of cognitive processing. This, and many other supportive findings, tip the balance clearly in favor of an affective explanation.
8 A similar experiment was conducted by Monahan, Murphy, and Zajonc (2000). They used subliminal exposures to further rule out the possibility of cognitive misattribution. In their study, 74 participants were exposed to a set of affectively neutral Chinese characters and were subsequently asked to report their mood. The experiment consisted of two conditions: in the repeated-exposure condition, participants viewed the same five characters for five times in a random sequence. In the single-exposure condition, participants viewed a series of 25 distinct characters. Participants then reported their mood by matching it to pictures of smiling faces, and completed two 5-point bipolar scales ranging from ‘happy’ to ‘sad’ and from ‘upbeat’ to ‘depressed’. Results showed that participants in the repeated-exposure condition selected more positive facial expressions and reported a better mood on both measures than participants in the single-exposure condition. Thus, repeated exposures of affectively neutral stimuli can induce a positive affective state. Cognitive theories of (mis)attribution are not supported, as the exposed stimuli could not be perceived consciously.
The most compelling evidence for a direct link between familiarity and positive affect probably comes from studies reversing the direction of this link. Garcia-Marques, Mackie, and Claypool (2004) found that positive affect leads to a feeling of familiarity. In their study, participants were explicitly exposed to 24 faces, of which 12 were smiling, and 12 were neutral. After an unrelated filler task, participants were exposed to the same set of faces again, now mixed with 24 new faces, of which 12 were smiling, and 12 were neutral. Participants had to judge for each face whether they had seen the face before (‘old’) or not (‘new’). Results showed that almost all familiar faces were correctly identified as ‘old’. However, quite a few novel faces were incorrectly identified as ‘old’ too. Interestingly, these false recognitions mainly consisted of smiling faces, as apposed to neutral ones. It can be concluded that a positive stimulus is more likely to be perceived as familiar than a neutral stimulus, supporting the notion of a bidirectional link between familiarity and positive affect.
9 A continuation of the effect of positive affect on perceived familiarity was found by Duke, Fiacconi, and Köhler (2014). They reported that primes with happy facial expressions led to an increase in perceived familiarity of subsequently presented memory probes. Further support for a bidirectional link was found by Monin (2003). In five experiments, results showed that positively-valenced stimuli are perceived as more familiar than neutral stimuli. However, it was unclear whether there is only a link between familiarity and positive affect, or whether novelty is linked to negative affect too.
With their correspondence hypothesis, Phaf and Rotteveel (2005) argued that this is indeed the case. They postulate that whether a stimulus is familiar or novel is processed at least as early as affect, dismissing attribution as an intervening process. They hypothesized that positive and familiar stimuli induce a mode of global processing, leading to less critical and more heuristic decision tendencies. On the other hand, novel and negative stimuli were expected to lead to analytic, critical and cautious processing. In a recognition memory test, 36 participants were exposed to previously-presented or new test words, followed by primes that were either matching, non-matching, positive, or negative. Indeed, positive affect led to a more liberal recognition bias, and negative affect led to more cautious tendencies.
To add weight to the correspondence hypothesis, a second experiment investigated whether the interaction between more diffuse mood states and familiarity was bidirectional too. Participants were asked to make old-new judgments of test words, while unconsciously altering their own affective states. Positive affect was induced by holding a coffee stirrer between the teeth (using the zygomaticus muscle) and negative affect was induced by furrowing the brow (using the corrugator muscle). In a control condition, participants were instructed to juggle a pen in the nonpreferred hand. Results showed that the positive manipulation resulted in a greater bias to judge neutral stimuli as old, as compared to the negative manipulation. Thus, even a very subtle mood manipulation can elicit a sense of
10 familiarity, implying that the link between memory and affect is very intimate.
Sufficient evidence now seems available to reject the misattribution models. Two remaining affective theories comprise the hedonic fluency hypothesis (Winkielman, Schwarz, Fazendeiro, & Reber, 2003) and the affective monitoring hypothesis (Phaf & Rotteveel, 2012).
The hedonic fluency hypothesis posits that the fluency signal is ‘hedonically marked’: it is believed to directly and inherently evoke positive affect. Several studies that manipulated processing fluency in the form of perceptual fluency (e.g. by altering figure-to-ground
contrast (Reber, Winkielman, & Schwarz, 1998)); semantic fluency (e.g. by using congruent instead of discongruent word combinations (Jacoby & Whitehouse, 1989)); and conceptual fluency (e.g. by varying direction of arrows in reading direction (Phaf & Rotteveel, 2009), or time duration, (Vliek & Rotteveel, 2012)) support this theory. These manipulations
consistently enhanced positive evaluations.
For example, Winkielman et. al (2001) found a link between processing ease and zygomaticus region activity. In this study, participants were exposed to a series of neutral pictures of everyday objects (e.g. a house, dog, or bird) and after each exposure, they reported how much they liked the object. Fluency was manipulated by a subliminally presented prime that was flashed prior to the target and either matched it or not. Affect was measured by facial EMG of the zygomatic and corrugator muscles. Results showed more activation of the
zygomaticus muscle when pictures followed matching primes (fluent) than when the the picture did not follow a matching prime (disfluent). However, no relation between disfluent primes and negative affect was found. This is in line with the hedonic fluency hypothesis, that postulates a selectively positive effect of fluency on affect.
In a second experiment, these results were extended by altering the nature of the fluency manipulation. Again, participants watched pictures of everyday objects. This time,
11 processing fluency was manipulated by varying the presentation duration of the pictures. Longer presentation durations were expected to increase processing fluency and thus lead to more positive affect than shorter presentation durations. Affect was again measured by EMG and self-reports. Results indicated that longer presentation durations were indeed associated with more positive affect. It was concluded that processing ease elicits positive affect. Results of both experiments confirmed the predictions formed by the hedonic fluency hypothesis. In combination with other studies reporting positive effects of enhanced fluency in a number of different ways, it seems plausible that fluency is able to directly elicit positive affect.
Despite its apparent strong foundation, the hedonic fluency model has two essential shortcomings. First, the distinction between familiarity and fluency is not necessary (Phaf & Rotteveel, 2005). Instead, it can be argued that familiarity and fluency merely reflect a difference in name but not in process. If anything, familiarity is a form of fluency, but both essentially represent the absence of neural competition. Both familiarity and other types of fluency can directly elicit positive affect, but do not seem to do so under all circumstances (Phaf & Rotteveel, 2012). This brings us to the second shortcoming of the hedonic fluency hypothesis: it does not account for the finding that continuous fluency will eventually lead to boredom and to negative affect. Over time, positive evaluations of repeatedly exposed stimuli have been found to decline and will even elicit a negative reaction (Bornstein et al.,1990). The present findings call for a model that recognizes the affective nature of familiarity and novelty processing, while retaining that fluency does not unconditionally lead to positive affect. Furthermore, for fluency signals to provide feedback on ongoing inormation
processing, they would have to act in parallel with the process they are monitoring. Thus, an appropriate model should describe a mechanism that is able to monitor its own dynamics. Such a basic monitoring mechanism has been proposed by Phaf and Rotteveel (2012). Their affective monitoring hypothesis describes a basic mechanism that detects changes on a
12 neural level, and is regarded as one of the most elementary operations performed by the brain. When a stimulus is encountered that has an existing representation in memory (as is the case with a familiar stimulus), a match will occur and processing of the stimulus will be fluent. However, when a novel stimulus is encountered, no existing representation can be found and a mismatch will occur, leading to disfluent processing of the stimulus. While negative affect is directly elicited by a conflict, mismatch, or disfluency, positive affect is only elicited when such a conflict can be resolved quickly by a unequivocal response, match, or fluent
processing. In other words: affective monitoring does not argue that fluent processing will intrinsically evoke positive affect, but emphasizes that a prior conflict is necessary in order for positive affect to be elicited.
That initial conflict is needed for the elicitation of positive affect, was already
suggested by the tension-release hypothesis of Sroufe and Waters (1976). They hypothesized that tension is necessary for the elicitation of laughter. As tension is a state of uncertainty and conflict, it corresponds to negative affect. Tension is first raised by uncertainty or ambiguity and its subsequent fast reduction leads to smiles and laughter, as a smile entails the relief from built-up tension. Tension release is also the essence of a good joke, tension is first raised and then quickly resolved. The tension-release hypothesis further supports the idea that initial competition is needed in order to evoke positive affect.
More support for the affective monitoring hypothesis comes from Topolinski and Reber’s (2010) analysis of the “Aha!”-experience. Also known as the ‘eureka effect’ or simply as insight, it refers to the common human experience of suddenly understanding something that was previously incomprehensible. Topolinski and Reber posit that the resolution of such initial conflict is what elicits positive affect. Consider the following example: a student has been doing her mathematics homework all afternoon, but does not seem to be able to fully grasp the theory. At this point, she is experiencing neural conflict,
13 uncertainty, and negative affect. Suddenly, however, an unexpected insight makes that the formula is somehow grasped: in a rush, the “Aha!”-feeling is experienced. The student is now overwhelmed with positive affect. The fact that this resolution is sudden and unexpected, and happens without previous awareness of the person experiencing it, ever more strongly
dismisses the possibility that the effect is caused by an indirect cognitive misattribution. Here too, the sudden onset of fluency coincides with the elicitation of positive affect.
As noted before, affective monitoring further argues that neural competition directly elicits negative affect. Support for this notion has been found by Dreisbach & Fischer (2012). In an affective priming paradigm, participants were primed with a Stroop-stimulus, and then exposed to a picture or a word, which they had to label as ‘positive’ or as ‘negative’.
Cognitive conflict was manipulated by either a congruent prime (colour matches colour meaning of the word), or an incongruent prime (colour mismatches colour meaning). The subsequent targets were either positively or negatively valenced. Results showed that negative targets were evaluated faster after incongruent primes than after congruent primes. Likewise, positive targets were evaluated slower after incongruent primes than after congruent primes. It was concluded that conflict is automatically associated with negative affect, which in turn leads to faster processing of negative targets and a slower processing of positive targets. So far, the two basic premises of affective monitoring have been supported, i.e. 1) novelty is primarily negative, and 2) when conflict is quickly followed by fluent processing, positive affect will be elicited.
In conclusion, familiarity is primarily associated with positive affect, whereas novelty corresponds to negative affect. Two-step theories are contradicted by a body of research indicating that genuine positive affect is elicited without the intervention of attribution. Hedonic fluency theories are not supported due to the finding that fluency simply does not always elicit positive affect. A basic monitoring mechanism in the form of affective
14 monitoring offers the most complete and straightforward explanation of current findings. Within a possible dual route mechanism for affect elicitation, affective monitoring therefore seems the most befitting model constituting the direct pathway.
In the next section, the link between novelty and positive affect will be discussed. If novelty is equal to neural conflict and direct negative affect, can it still evoke positive affect?
Can Novelty Be Positive?
Although the link between familiarity and positive affect is strongly supported, it is tenable that humans also hold a natural tendency to explore what is novel. In this section, I will discuss how boredom effects, arousal, and even variation in mood and mindset can lead to a preference for novelty. These findings will then be viewed in the light of evolutionary prepared behavior systems, that act in parallel to affective monitoring.
A preference for novel over familiar stimuli has previously been evidenced by Berlyne (1970), who argued that boredom limits the mere exposure effect. Similar findings were presented by Stang (1975), who suggested that the processing of novel stimuli is intrinsically rewarding. Furthermore, Bornstein et al. (1990) argued that we prefer complex, highly interpretable stimuli over familiar ones. After repeated exposures, they found preferences for stimple stimuli to decline more than preferences for more complex, optical illusions.
Additionally, only participants with low boredom-proneness showed a mere exposure effect. These results might lead to the idea that novelty is inherently preferred over familiarity. However, this conclusion is premature and undue.
The notion of boredom can be explained by affective monitoring. When processing is fluent for a long time, boredom will arise. This is because the initial conflict has worn off and fluency no longer elicits positive affect. As was discussed in the previous section, not all fluency and familiarity lead to positive affect. Only when the fluency results from a resolution
15 of neural competition (i.e., a question, conflict, or problem), is positive affect elicited.
However, such an initial conflict has long worn off in the case of boredom. In this case, familiarity may lead to negative affect and a bias towards novelty preference arises. Boredom is not the only factor that can motivate a preference for novelty. For
example, Harmon-Jones and Allen (2001) found that individual differences in affective styles moderate familiarity preference. In their experiment (earlier described in the previous
section), they additionally found that the mere exposure effect was strongest in participants with the lowest baseline of self-reported affect. Thus, familiarity preference is highest when positive affect is low. This finding conforms to the idea of familiarity as a safety cue
(Bornstein, 1989). Assuming that positive affect signals a safe environment, and negative affect signals an unsafe environment, familiarity should be particularly valued in a negatively valenced affective state (De Vries, Holland, Chenier, Starr, & Winkielman, 2010).
Correspondingly, familiarity should be less valued in a benign environment or in a state that is already affectively positive.
Such effects of mood on reduced familiarity preference have been the subject of a study by De Vries et al. (2010). They instructed participants to write a happy or a sad autobiographical story in order to manipulate mood. Then, participants rated random dot patterns that were either familiarized or novel. Results showed that sad participants preferred familiar patterns, while happiness eliminated this preference. The notion is thus supported that a positive mood entails safety, and gives room to exploratory tendencies, such as novelty seeking.
Building on this, optimal level of arousal (OLA) theories argue that deviations of one's personal optimal level of arousal can motivate behavior designed to attain it. The basic
premise is that someone ‘under’ their optimal level of arousal will try to increase arousal by seeking novel stimulation. Identically, someone ‘above’ their optimal level of arousal will try
16 to decrease arousal by keeping to what is (safely) familiar. Pliner and Loewen (2002)
examined the effects of manipulated arousal on children’s willingness to try novel foods. In their study, 165 children were assigned to either a high-, moderate-, or low-arousal condition. Arousal varied between conditions in presented video clips, music, and the games the children got to play. Results showed that children in a low arousal condition were willing to try more novel foods than children in a high arousal condition. It can be concluded that a lowered level of arousal may motivate novelty seeking behavior.
The aforementioned studies all seem to indicate that novelty is particularly preferred when one’s current state is positive, safe, and calm. Even though novelty may fundamentally elicit neural competition, this does not always seem to result in negative affect. However, it is hard to tell whether novelty truly elicits positive affect in these studies, or whether the results merely reflect a decrease in familiarity preference. For example, as in the study by De Vries et al. (2010) it could be argued that a happy mood sets up expectations of high fluency, which prevents the elicitation of any effect. Remember that in the light of affective monitoring, positive affect is only elicited when it is preceded by conflict. When someone is already in a happy mood, this conflict may be absent and neither novelty or familiarity will evoke any affect (Phaf & Rotteveel, 2012).
As for the notion of an ‘optimal’ level of arousal, it must be acknowledged that this concept is hard to define. Perhaps it can better be explained in terms of neural conflict. As high arousal stands for a high baseline level of conflict, low arousal corresponds to a low level of conflict. By redefining this definition, findings within the OLA paradigm can be
consolidated within the affective monitoring hypothesis.
As mentioned before, we might be evolutionarily predisposed to encounter something familiar with less caution than something novel. For this reason, different modes of
17 prepared behavior systems may modulate a bias towards familiarity or novelty preferences. In this light, Panksepp (1998) distinguished a set of behavior systems, of which ‘SEEKING’ and perhaps ‘PLAY’ encompass the most relevant ones to novelty preferences and exploration. It is hypothesized that it is monitored whether the ongoing activity matches the goal of the activated behavior system. If this is the case, positive affect will be elicited. If not, a conflict takes place and negative affect will arise. In this manner, novelty can become a sign of fluent processing and indirectly evoke positive affect. For example, when the ‘SEEKING’ system is activated and something novel is found, this will lead to positive affect because it is in
accordance with the goals of the currently activated system.
If such behavior systems can account for variation in novelty/familiarity preference, it can be hypothesised that activation of the tendencies involved in such a behavior system through for instance mindset or mood, may induce a preference for either novelty or familiarity.
Hansen and Topolinski (2011) argued that an exploratory mindset increases liking for novel objects. 54 participants were asked to judge the attractiveness of dot patterns. Patterns consisted of prototypes of familiar categories, exemplars of familiar categories, or exemplars of novel categories. Mindset was manipulated by asking participants to imagine the stimuli as stars in the exploration condition, or as peas in the control condition. They hypothesized that in contrast to pea configurations, star constellations would induce a stronger exploratory mindset as they are related to both mythology and scientific exploration. Results showed that participants in the exploration condition preferred novel exemplars, while participants in the control condition preferred prototypes. It can be concluded that mindset can modulate a preference for novelty. An exploratory mindset can be directly traced back to activation of an exploratory behavior system, such as SEEKING.Furthermore, Topolinski and Sparenberg (2012) found that rotating a handle clockwise (as a universal symbol for progression towards
18 the future and novelty) enhances novelty preference and that turning it counterclockwise (as a universal symbol for regression and going back to what it old) enhances familiarity
preference. By rotating a handle clockwise, a tendency to explore is activated, motivating behavior towards this goal.
In this section, studies reporting a preference for novelty over familiarity have been presented. While intuitively contrary to affective monitoring, this novelty preference can be explained in the light of activation of different behavior systems. Alternatively, it could be argued that some of the presented findings result from a different mode of processing, e.g. as proposed by a dual-route framework. In the next section, a dual route framework will be provided that might shed more light on when familiarity, and when novelty may elicit positive affect.
A Dual-Route Framework
As was suggested in the introduction of this thesis, it is possible that affect elicitation from novel and familiar stimuli can be interpreted in the light of a dual-route model. In such a model, stimuli are subject to both direct and indirect routes of processing. A direct route consists of global processing and is assumed to be nonconscious. Conversely, an indirect route consists of detailed or analytical processing and is assumed to be more conscious. Furthermore, these processes are assumed to take place in parallel (Phaf & Rotteveel, 2005). In this section, I will argue that such a dual route model is plausible.
The notion of a dual route model has already been proposed by the dual-process theory of Oatley and Johnson-Laird (1987). Here, two different signals are distinguished: the
‘cognitive’ informational signal and the ‘affective’ control signal. The former provides information about the particular cause of an emotion and to what or whom the emotion is directed. The latter is quick, relies on heuristic cues, and has an emotional tone. Similarly,
19 LeDoux (1996) distinguished a direct and an indirect route, emphasizing that the direct route entails quick, global processing and the indirect route allows for a more detailed analysis of a stimulus. Although nonconsciously perceived objects can only be processed via the direct route, consciously perceived objects can be processed by both. In this light, it makes sense that the same stimulus can elicit positive affect when it is presented suboptimally (e.g.
because it is familiar), but evokes negative affect when it is presented optimally (e.g. because it is a picture of something you personally do not like, and this inhibits the originally evoked positive affect).
According to Schwarz (2010) processing fluency can indeed lead to evaluative
judgments via two different routes. The first route consists of the spontaneous, positive affect that is elicited by processing fluency. Fluency effects via the second route are subject to how the fluency is interpreted. Jian and Hong (2014) named the first route the ‘affective pathway’ and the second route the ‘interpretive’ pathway. Whereas with the affective pathway fluency leads to positive affect, processing through the interpretive pathway can cause fluency to be evaluated either as positive or as negative, depending on contextual factors.
Jian and Hong (2014) found that the extent of cognitive load determines what route we use to process information. When enough cognitive load is available, we have the resources to process information via the interpretive pathway. However, when cognitive load is increased, we will only be able to rely on the affective pathway, which is more automatic and requires less mental resources. In their experiment, Jian and Hong built on the finding that when expectancies (e.g. of time duration) are violated, affective judgment will tend to be negative as processing fluency is decreased (Vliek & Rotteveel, 2012). They instructed participants to evaluate a fictitious brand of medicine (‘Lipiter’). Prior to the experiment, half of the
participants was instructed to remember a 13-digit number (increasing cognitive load) that would need to be recalled afterwards. Expected fluency was manipulated by varying the
20 product category. In high-expected-fluency conditions, Lipiter was a simple cold medicine. In low-expected-fluency conditions, Lipiter was an anticoagulant medicine. Experienced
processing fluency was manipulated by altering the font of the advertisement. In high-experienced-fluency conditions Times New Roman was used. In low-high-experienced-fluency conditions, the harder to read Script MT Bold was used. In the congruent condition, expected fluency matched experienced fluency. In the incongruent condition, expected fluency did not match eperienced fluency. Participants evaluated the product using two scales, ranging from ‘bad-good’, and from ‘ineffective-effective’. Results showed that in congruent conditions, brand evaluations were more positive than in incongruent conditions. However, when cognitive load was increased, participants only preferred items that were processed fluently, regardless of expectancies. This indicates that there is a dissociation between a direct (affective) route, where the experience of fluency elicits positive affect, and an indirect (interpretive) route, where evaluations are more malleable and subject to contextual interpretations, giving room to either familiarity or novelty preferences.
While further research is needed to shed more light on when each route comes into play, it is likely that novelty preferences occur at a later stage and are different and
independent from the immediate positive affect that is elicited by fluency at an earlier stage.
Conclusions and Discussion
Memory processing and affect are intimately linked. Something that is familiar to us can directly elicit genuine positive affect (Harmon-Jones & Allen, 2001), and something we perceive as positive will elicit a feeling of familiarity (Monin, 2003). Not only does
familiarity correspond to positive affect; novelty and negative affect are analogously linked (Phaf & Rotteveel, 2005). A basic model that illustrates how affect can be elicited from neutral stimuli is proposed in the form of affective monitoring (Phaf & Rotteveel, 2012).
21 Affective monitoring can account for experimental findings of affect elicitation where other theories, such as the hedonic fluency model (Winkielman & Cacioppo, 2001) fall short. According to affective monitoring, novelty is linked to negative affect, because novelty elicits neural competition. The swift resolution of this competition will elicit positive affect, as competition is resolved quickly by fluent processing. The notion that initial conflict is needed for affect to be elicited, is supported by the tension-release hypothesis (Sroufe & Waters, 1976) and accounts of the universally common “aha!”-experience (Topolinski & Reber, 2010), where the release of built-up tension results in positive affect.
Indirectly, however, both novelty and familiarity can induce either positive or negative affect. Novelty may primarily be positive as a relief from boredom (Bornstein et al., 1990). Additionally, novelty preference can be explained in light of behavior systems (Panksepp, 1998). For example, when a behavior system designed to seek novelty is activated, an encounter with a novel stimulus will elicit positive affect (Phaf & Rotteveel, 2012). Differences in mood (De Vries et al., 2010) or mindset (Hansen & Topolinski, 2011) can modulate novelty preference, which can be partially explained by the notion that familiarity serves as a cue to safety, and that exploratory behaviour is not inhibited when an individual is in a safe and familiar environment (De Vries et. al, 2010). When something is perceived as novel, affective state will likely be negative, and we will be more concerned with behavior towards safety. Conversely, when we perceive something as familiar, affective state will likely be positive and we will be less concerned with danger and more with exploration. The paradox herein lies that the more familiarity we experience, the more open we are to novelty. And the more novelty we experience, the more we seem to hold on to what is familiar. Behavior systems operate in parallel to affective monitoring, and it is the latter that imparts an affective ‘flavor’ to the former (Phaf & Rotteveel, 2012). Both systems operate at a very early stage of information processing, and their basic affective reactions could be
22 inhibited by a more thorough mode of processing that allows information to be evaluated via an interpretive pathway as opposed to an early, affective pathway (LeDoux, 1996; Jian & Hong, 2014). In the light of such a dual route framework, seemingly contradicting findings concerning novelty and familiarity preference can be explained by this dissociation in a direct and an indirect mode of processing.
Gaining insight into the link between information processing and affect has important implications for current scientific research. As memory and affect have traditionally been regarded as two predominantly separate subjects, the mere notion that affect can be derived from processing dynamics itself changes the way we perceive their relation to one another. For example, Garcia-Marques et al. (2010) noted that “Like Moliére's Bourgeois Gentleman, Monsieur Jordain, who one day realized that he had unknowingly been speaking prose all his life, perhaps we should also begin to realize that by studying memory, we are also studying affect.” (p. 216). As the gap between memory and affect becomes increasingly smaller, a vast amount of valuable knowledge will become available to scientists of both fields.
Furthermore, research in this area is relevant for many real-world situations. For example, a company might deem a familiar candidate with a smile on his face a better option than a new candidate with a smile on his face. In addition, one particular CV might directly appeal to a manager because it is more fluently processed than the others, e.g. because the font or layout are better to read or more familiar than those used in other CVs. In addition, a company that has been through a rough time with many changes and great periods of
insecurity, may likely opt for familiar, conventional candidates – mirroring the notion of familiarity as a safety cue. Conversely, an unconventional candidate or CV might be desired when the assessor is in a novelty seeking mindset, e.g. when he is bored and all CV’s so far looked the same.
23 suggestions for further research. Consider the example that you need to decide whether you recognize a suspect or not. When you are in a positive mood, you will be more likely to believe you have seen him before.
Another scope for future research would be to clarify the mechanism of how affect can be derived from neutral stimuli, as is central to affective monitoring. Furthermore, while a direct and an indirecte route of processing can be distinguished, it is still unclear which route will be activated in what situation. Even though a basic mechanism has been described, a person may rely on several cues (i.e. familiarity as a safety cue) that influence preference for novelty or for familiarity. In uncovering factors that modulate these preferences, it will become increasingly realistic to predict whether novelty or familiarity is desirable in any given situation.
And what about Dorothy? She would probably show a normal mere exposure effect. However, in terms of behavior systems, her exploratory tendencies would probably take over more readily than, say, Karl’s. In terms derived from optimal level of arousal theories (Pliner & Loewen, 2002), Dorothy’s baseline level of conflict is presumably lower than Karl’s, explaining her hunger for (and his aversion of) novel experiences.
In a way, we are all a bit like Dorothy. We might not celebrate our lives each year with the thrill of risking to lose it, but we certainly experience novelty in our own ways. Some might feel adventurous and pick up a book from an unknown author at the local bookstore, while others might be so prone to boredom that they can’t finish a book without starting another three. While we may be wired to like the familiar, our selfish genes will continue to drive us to explore the novel.
24 References
Berlyne, D. E. (1970). Novelty, complexity, and hedonic value. Perception & Psychophysics, 8(5), 279-286.
Bornstein, R. F. (1989). Exposure and affect: Overview and meta-analysis of research, 1968– 1987. Psychological Bulletin, 106, 265-289.
Bornstein, R. F., Kale, A. R., & Cornell, K. R. (1990). Boredom as a limiting condition on the mere exposure effect. Journal of Personality and Social Psychology, 58, 791-800. Claypool, H. M., Hall, C. E., Mackie, D. M., & Garcia-Marques, T. (2008). Positive mood,
attribution, and the illusion of familiarity. Journal of Experimental Social Psychology, 44, 721-728.
de Vries, M., Holland, R. W., Chenier, T., Starr, M. J., & Winkielman, P. (2010). Happiness cools the warm glow of familiarity: Psychophysiological evidence that mood
modulates the familiarity-affect link. Psychological Science, 21, 321-328.
Dreisbach, G., & Fischer, R. (2012). Conflicts as aversive signals. Brain and Cognition, 78, 94-98.
Duke, D., Fiacconi, C. M., & Köhler, S. (2014). Parallel effects of processing fluency and positive affect on familiarity-based recognition decisions for faces. Frontiers in Psychology, 5:328 10.3389/fpsyg.2014.00328.
Garcia-Marques, T., Mackie, D. M., Claypool, H. M., & Garcia-Marques, L. (2004). Positivity can cue familiarity. Personality and Social Psychology Bulletin, 30, 585-593.
Hansen, J. & Topolinski, S. (2011). An exploratory mindset reduces preference for prototypes and increases preference for novel exemplars. Cognition and Emotion, 25, 709-716.
25 Harmon-Jones, E., & Allen, J. J. (2001). The role of affect in the mere exposure effect:
Evidence from psychophysiological and individual differences
approaches. Personality and Social Psychology Bulletin, 27, 889-898.
Jacoby, L. L., & Whitehouse, K. (1989). An illusion of memory: false recognition influenced by unconscious perception. Journal of Experimental Psychology - General, 118, 126-135.
Jiang, Y., & Hong, J. (2014). It feels fluent, but not right: The interactive effect of expected and experienced processing fluency on evaluative judgment. Journal of Experimental Social Psychology, 54, 147-152.
Larsen, J. T., Norris, C. J., & Cacioppo, J. T. (2003). Effects of positive and negative affect on electromyographic activity over zygomaticus major and corrugator
supercilii. Psychophysiology, 40, 776-785.
LeDoux, J. (1996). The emotional brain: The mysterious underpinnings of emotional life. Simon and Schuster.
Mandler, G., Nakamura, Y., & Van Zandt, B. J. (1987). Nonspecific effects of exposure on stimuli that cannot be recognized. Journal of Experimental Psychology: Learning, Memory, and Cognition, 13, 646.
Monahan, J. L., Murphy, S. T., & Zajonc, R. B. (2000). Subliminal mere exposure: Specific, general, and diffuse effects. Psychological Science, 11, 462-466.
Monin, B. (2003). The warm glow heuristic: when liking leads to familiarity. Journal of Personality and Social Psychology, 85, 1035-1048.
Oatley, K., & Johnson-Laird, P. N. (1987). Towards a cognitive theory of emotions. Cognition and emotion, 1, 29-50.
Panksepp, J. (1998). Affective neuroscience: The foundations of human and animal emotions. Oxford university press.
26 Phaf, R. H., & Rotteveel, M. (2005). Affective modulation of recognition bias. Emotion, 5,
309-318.
Phaf, R. H., & Rotteveel, M. (2009). Looking at the bright side: The affective monitoring of direction. Emotion, 9, 729-733.
Phaf, R. H., & Rotteveel, M. (2012). Affective monitoring: A generic mechanism for affect elicitation. Frontiers in Psychology, 3. http://dx.doi.org/10.3389/fpsyg.2012.00047. Pliner, P., & Loewen, R. (2002). The effects of manipulated arousal on children's willingness
to taste novel foods. Physiology & Behavior, 76, 551-558.
Reber, R., Winkielman, P., & Schwarz, N. (1998). Effects of perceptual fluency on affective judgments. Psychological science, 9, 45-48.
Schwarz, N. (2010). Meaning in context: Metacognitive experiences. In B. Mesquita, L. F. Barrett, & E. R. Smith (Eds.), The Mind in Context (pp. 105–125). New York, NY: Guilford.
Stang, D. J. (1975). Effects of "mere exposure" on learning and affect. Journal of Personality and Social psychology, 31, 7-13.
Sroufe, L. A., & Waters, E. (1976). The ontogenesis of smiling and laughter: a perspective on the organization of development in infancy. Psychological Review, 83, 173-189. Topolinski, S., & Reber, R. (2010). Gaining insight into the ‘‘aha’’experience. Current
Directions in Psychological Science, 19, 402–405.
Topolinski, S., & Sparenberg, P. (2012). Turning the hands of time clockwise movements increase preference for novelty. Social Psychological and Personality Science, 3, 308-314.
Vliek, M. L. W., & Rotteveel, M. (2012). If time flies, are you more fun? The relative effect of expected exposure duration on the evaluation of social stimuli. European Journal of Social Psychology, 42, 327–333.
27 Winkielman, P., & Cacioppo, J. T. (2001). Mind at ease puts a smile on the face:
Psychophysiological evidence that processing facilitation elicits positive affect. Journal of Personality and Social Psychology, 81, 989-1000.
Winkielman, P., Schwarz, N., Fazendeiro, T, & Reber, R. (2003). The hedonic marking of processing fluency: Implications for evaluative judgment. In J. Musch & K.C. Klauer (eds.), The psychology of evaluation: Affective processes in cognition and emotion (pp. 189-217).
Zajonc, R. B. (1968). Attitudinal effects of mere exposure. Journal of Personality and Social Psychology, 9, 1-27.
