Experimental Educational Psychology. expected level of sounds heard in a kitchen (Lemaitre et al., 2010). Journal of Experimental Psychology: Human Perception and Performance. Auditory semantic networks for words and natural. Our selection was guided by a list of sounds that people, usually hear in their kitchen. Different categories with clear, boundaries and structures allowed us to perform a textual data, analysis rather than an analysis of the verbalizations in combina-, tion with additive tree representations, which requires more inter-, pretation by the person analyzing the data. We also, observed a few terms related to acoustic properties and appraisal, the co-occurrences of terms in the descriptions, we observed that, the four main clusters of sounds (A, B, C, D) clearly overlapped, Developed by Image (http://www.image-zafar.com/english/index_, were associated with one CRV that corresponded to one homoge-, neous lexical field. By characterizing the impact of real-life transports on time perception, we will help understand subjective processes in decision-making for choosing between transport itineraries, in turn helping the elaboration of new tools to help navigation. We did not observe clusters of actions. Iterating, the second fitted matrix is now subtracted from the original, proximity matrix, and a first (re)fitted matrix is obtained. The Representation of Proximity Matrices by Tree Structures: 5. (2007) claimed that this map was incom-, plete because several categories are missing (vocalizations and elec-, tronically synthesized sounds). These, three levels of relation (acoustical, physical production of the, sounds, semantic descriptions) can be a powerful tool for searching, Abdi, H. (2007). The International Journal of Psychology (IJP) advances psychological research of interest and relevance for the human condition around the world and serves as the outlet for internationally relevant basic and applied research in scientific psychology.. Across these two experiments, our aim was as follows: • To observe whether the classes created by the participants, reflected different categories of physical events, as hypothesized. He proposed that, properties of a sound (i.e., the listener describes perceptual qual-, ities, such as pitch and timbre). The corpus was divided into 117, ECUs, and 90% of the ECUs were taken into account during the, that the verbs used to describe the sounds clearly described the, gories of representative verbalizations CRV. (1977). The, goal of these experiments was not to test the internal structure of, the category (e.g., typicality between sounds within a category) or, to test different theoretical approaches (e.g., comparing the proto-, typical view and the exemplar view). It also uses basic naviga-, tion processes based on the semantic descriptions of sounds in, commercial sound databases such as Sound Ideas (http://, www.sound-ideas.com/). Read the latest issue and learn how to publish your work in Educational Psychology. Primary dendrogram of the hierarchical cluster analysis (Hubert, Arabie, & Meulman, Experiment 2. and Applied Mathematics, American Statistical Association. Figure 6 represents the 13, coefficients are computed for each pair of individual proximity matrices of the participants and calculated, distances between the results of participant classifications. This procedure involves the adjustment of the coefficients of the first four parameters as well as the eventual introduction of new ones to attain a perceptually relevant distance between two musical sounds. For each of these basic events, the author suggested a set of, physical properties that may be perceptually relevant or may, influence perception (e.g., for the sounds of scraping, the physical, The author also assumed that different combinations of basic, events produce more complex events. This is a major difference from Gaver’s (1993b) view. To summarize, we identified five CRVs. Journal of Experimental Psychology: Learning, Memory, and Cognition What Limits Working Memory Capacity? For each of these, a transformation factor between the physical axis and the corresponding subjective dimension is obtained by linear regression. The effect of stimulus group and methodology. As shown by the review of classification studies, a hierarchical, structure is probably too simple to account for the human catego-, rization of sounds. The action of falling precedes the action of rolling. Journal of the Acoustical Society of America. As in the. Hubert, L. J., Arabie, P., & Meulman, J. secondary dendrogram is reported in Appendix C. Figure 3 and Appendix C represent the primary and the second-, ary dendrograms, respectively. Approach to Auditory Event Perception, Copyright 1993 by Taylor &. A listener can use both modes of, listening for the same sound: “The distinction between everyday, listening and musical listening is between experiences, not sounds”, In general, the literature suggests that listeners tend to favor, everyday listening and focus on the physical phenomena that cause. The convergence between the structure of the clusters, (hierarchical cluster analysis of the sorting tasks) and the structure, of the lexical categories (lexical analysis of the verbalizations) can, be understood by two assumptions. Appendix Table H1 re-, To summarize, the 13 clusters were described mainly by the, identification of the physical action producing the sounds. The different clusters reflected categorization at the level of the, action or movement generating sound. They. First lexical analysis: Categories of representative verbaliza-, software; Reinert, 1986) to determine whether the lexical fields, extracted from the verbalizations could reflect how the main, clusters (A, B, C, and D) were structured. We analyzed the interindividual differ-, ences between the individual proximity matrices of the 30 partic-, ipants. Categories are, dependent on the context of the task. (1997) required their subjects to describe their cate-. Sound indexing using morphological. (2009). teria as in the first experimental study. & Collins, 1993). Hear? chi-square, indicating the importance of the words in the CRV. Journal of Experimental Psychology: Human Perception and Performance When Less Is More: Impact of Face Processing Ability on Recognition of Visually Degraded Faces Jessica Royer, Caroline Blais, Fr d ric Gosselin, Justin Duncan, and Daniel Fiset Balls dropped on plates. mainly based on the identification of the sound sources (solid, liquid, gas, and electric devices), similar to the analysis of the main, clusters. The previously presented experimental studies used dif-, ferent types of instructions, and their results were examined with, different statistical and lexical analyses. In general, the, descriptions appeared to be coherent with the labels of the sound, To summarize, the main Cluster A included sounds made by, solid objects and their associated actions (knife and cutting in A, grouped more abstract functions such as opening a can or a, packaging that involved an analogous goal, if not necessarily the, same actions, or produced a similar sound due to the manipulation, mix a script representation (i.e., a representation of a situation, organized around personal experiences in which human interaction, is central; Schank & Abelson, 1977), preparation of meal. Development of the database for environ-. Roma, G., Janer, J., Kersten, S., Schirosa, M., Herrera, P., & Serra, X. Vanderveer’s (1979) seminal work showed that participants, grouped sounds together that were caused by the same event, (dropping a pen, a can, a piece of wood) or that shared the same, acoustic properties. The corpus was divided into 96 ECUs, and 72% of the, ECUs were considered in the analysis. Anti-Robinson matrices for symmetric proximity data 10. These, sounds were simply identified as liquids. This work was also supported in part by National Science, Foundation 0946550 to Laurie Heller. The, sounds of liquids seemed to have a specific perceptual status: they, were systematically categorized as liquid sounds even if they could. When, a new sound is imported into the database, the sound is compared, with other sounds based on acoustical similarity. Hierarchical Cluster Analysis With Several Dendrograms, For the hierarchical cluster analyses, we used the freely, available algorithms provided at http://cda.psych.uiuc.edu/, srpm_mfiles/ (Hubert et al., 2006). They were allowed to form as many classes as they, wished and to put as many sounds in each class as they desired. classification and category organization in a complex real-world domain. Canadian Journal of Experimental Psychology/Revue canadienne de psychologie expérimentale Construction, Integration, and Mind Wandering in Reading Peter Dixon and Marisa Bortolussi Online First Publication, February 11, 2013. doi: 10.1037/a0031234 CITATION Dixon, P., & Bortolussi, M. (2013, February 11). on the sounds of solid objects (Experiment 2). They were selected using the same cri-. (2010). The. Auditory perception of. These, specific contexts may limit the formation of large semantic cate-, gories (e.g., the transport category) or more conceptual categories, (, e.g., the leisure category) involving semantic relations (, Wordnet network (Fellbaum, 1998). erties), the actions that produced the sounds (opening/closing, compressing, cutting, rubbing), the context (to cook) and a few, of objects (flexible vs. rigid) that constrained different types of, chines (appliances) and to appraisal judgments of the sounds, but. The Secret to Happiness: Feeling Good or Feeling Right? The sound designer can navigate in, this 2-D space in which reference sounds are located and may, choose a location to generate a new sound with new synthesis, parameters. In a previous study, Lemaitre, Houix, Misdariis, and Susini (2010) identified three. However, this approach only highlighted, general categories. The lexical forms were sorted into five semantic fields. The Journal of Experimental Social Psychology is devoted to the publication of theoretically grounded experimental research on social interaction and related phenomena, including attitudes; social cognition, perception, emotions and motivation; interpersonal relationships; social communication; and intragroup and intergroup processes. For Shakespeare’s, ysis can determine whether the different CRVs (the lexical field of, the action or related to the concept of possession, to use examples, from the tutorial) are related to specific act(s) or specific charac-, ter(s) (for example, the King and the Ghost, two characters in the, Experiment 1. This was probably be-, cause the other categories do not usually occur in a kitchen. Ecological acoustics perspective for content-based retrieval of, Rosch, E. (1978). These types of listening are illus-, trated by listeners’ reports of many different things when they, describe what they hear. Several studies dealing with the perception of musical timbre have found significant correlations between acoustical parameters of sounds and their subjective dimensions. interpret information about the environment and to interact with it. In the category of sounds, generated by solid interactions, the most general level is related to, the movements or actions generating sounds, implying a different, Experiment 2: Occurrences of the Lexical Forms in the Five Semantic Fields for the, Verbalizations of Each Main Cluster of the Primary Dendrogram (Figure 6), type of temporal patterning. sources (animals/people, vehicles/mechanical, musical and water). We, coefficient (Abdi, 2007) to calculate the agreement, between participants. He introduced different types of listening that imply an active role, and involve different levels of attention: unexpected and diverted, (with a global scope as a general scan of the environment) or, focused on a particular sound. Each CRV was therefore charac-. Using the conclusions of some of these studies, a calculation method of the perceptual distance between two sounds has been developed. In cognitive neuroscience, the notion of cognitive map is central for the spatial and temporal mapping of the environment: cognitive maps are a system of representations in the brain enabling an individual to infer its position in the environment. The verbalizations were produced in French following the sort-, ing task. For example, of meanings: the musical attributes or the musical instrument made, of different parts (a keyboard, a pedal, etc.). Overall, these analyses showed that the participants used, different criteria to sort the sounds. For example, Gygi and Shafiro, (2010) proposed an automatic classification of environmental, sounds related to the work of Gaver (1993b). In Experiment 3, we found that the structural invariant was reliably detectable but that the degree to which structural properties were preserved depended on the kind and amount of transformation applied. A low value of 0 indicates a perfect identification of the sound, and a value. The methodological aspects were important for, structuring our experimental studies, but our specific focus on, sound event categories forced us to identify criteria for selecting, This section presents a taxonomy of everyday sounds that rep-. Journal of Experimental Psychology: General VOL. The sounds seemed to be categorized as sound sources, Finally, we observed several other criteria for classifying the, sounds: acoustical properties, appraisal judgments, and more ab-, stract representations (function, script representation, etc.). An Ecological, Etude de la perception sonore en termes de reconnais-. The struc-, ture of the semantic categories overlapped the organization of the, main clusters of the primary dendrogram (Experiment 2), indicat-, ing a correspondence between the verbalizations and the catego-, different categories of representative verbalizations (CRVs; first analysis) and the different main categories and, their semantic portraits (second analysis) is depicted with lines. Another proposed approach is a struc-, ture for organizing sound databases that reflects how human lis-, teners categorize sound events. the same structure, such as the organization proposed by Guyot et, In the first experiment, lay participants were asked to freely sort a. set of environmental sounds with different degrees of identifiability. Similar, to Gaver’s (1993b) taxonomy, we found a compound category, mixing multiple impacts and deformation. and winds (continuous introductions of pressure variations). It includes empirical psychological research on cognitive , social / emotional , and physical development. The variance accounted for (VAF), had a value of .92, indicating that the multiple tree representation, To identify significant clusters in a dendrogram, it is usually cut, at a given fusion level. Nevertheless, we included two sounds of persons, walking on different floors (gravel and rubber), without a, specific gait. —Roger C. Schank and Robert P. Abelson from the Introduction (http://www.psypress.com/scripts-plans-goals-and-understanding-9780898591385). goal of this analysis was to produce a semantic portrait of each, subcluster of the primary dendrogram (Figure 3, Subclusters A, descriptions across participants for each subcluster. Their descriptions and classification appeared less reli-, able. The first assumption is a weak, influence of the language (French) on the categorization of these, sounds. As the title of the journal indicates, we are focused on publishing primary reports of research in social psychology that use experimental or quasi-experimental methods, although not every study in an article needs to be experimental. Eleven sounds were not. Nevertheless, this map describes, and analyzes a large part of the most common sound-producing, events, and it suggests the physical attributes that may be relevant, for a listener to identify these events. WildTimes will contribute to fundamental research by empirically contrasting theories of how the human brain represents time, and to practical applications with the conception and the optimization of novel tools to help individuals navigate transportation networks. categories (a situation exploited by Foley artists for sound design. Secondary dendrogram of the hierarchical cluster analysis (Hubert, Arabie, & Meulman, 2006), Experiment 2. clusters (1” to 13”) are indicated by Arabic numbers. [A study of sound perception, in terms of recognition and qualitative. However, the recovery of a sound event’s properties is not, always accurate and the recovery depends on whether sufficient, acoustic information is available and how efficiently listeners can, use it (Lutfi, 2001; Lutfi & Oh, 1997). This table shows, that the distribution of the lexical forms was primarily related to, the actions (61.16%) and was related to a lesser degree to the, objects (19.51%) producing the sounds. Therefore, some technologies use more com-, plex structures. (e7, e35, e17, e98, e101, e99, e38, e53, e97, cluster, the associated sounds are indicated (Appendix A). Complexity increases toward the center of the figure, with examples showing temporally patterned, compound, and hybrid sounds. 61 (Ultimate Sound Bank. ondary dendrogram is reported in Appendix G. There were 385 distinctive forms and 4,029 associated occur-, rences that were accounted as verbalizations produced by the, participants, and 228 analyzed forms and 3,581 occurrences that, were selected for the analysis. Journal of the Acoustical Society of America, 81, EURASIP Journal on Audio, Speech, and Music, Ecological acoustics: Human perception of, Hierarchical cluster analysis (Hubert, Arabie, & Meulman, 2006). An ecological, Giordano, B. L., & McAdams, S. (2006). (2001). These, actions were specific to each sound source. (sound events) grounded in the physical production of the sounds, which therefore constrains their acoustical properties. For this purpose, new acoustical descriptors related to the, spectro-temporal profiles of sounds should be developed, as in the, work by Peeters and Deruty (2010). Linear multidimensional scaling 3. Explorations in, Gaver, W. W. (1993b). inable entities as members of an equivalence class” (p. 246). As an alternative clustering method, in this, study, we used a threshold of inconsistency. Evidence for Modality-Specific Sources to the Simultaneous Storage of Visual and Auditory Arrays Daryl Fougnie and Ren Marois Online First Publication, August 22, 2011. doi: 10.1037/a0024834 CITATION Fougnie, D., & Marois, R. (2011, August 22). Thus, the second general lexical field appeared to, reflect actions or movements involving short, discrete interactions, In a second step, we analyzed the correspondence between the, CRVs and the clusters of sounds. Considering, a wide corpus of environmental sounds with a large variety of, categories (e.g., living and nonliving sounds, action and nonaction, sounds, and synthetic sounds that can imply a complex set of, semantic relations), an analysis with multiple dendrograms could, be useful to highlight complex relations between categories. (1997) authors proposed a framework for the, classification of environmental sounds based on these two modal-, ities. The Journal of Experimental Social Psychology (JESP) aims to publish articles that extend or create conceptual advances in social psychology. Read the latest articles of Journal of Experimental Social Psychology at ScienceDirect.com, Elsevier’s leading platform of peer-reviewed scholarly literature Obtaining the precise sound that the, sound designer or musician has in mind requires a mastery of the, underlying algorithms. The Roman numbers I, indicate the categories of representative verbalization CRVs identified by the analysis. erties that distinguished them from the other sounds (Gygi et al., Following the results of the first experiment, we focused on, the sounds produced exclusively by solid interactions. For, instance, we did not observe the categories whoosh, explosion, and, wind (Gaver, 1993b). First, we analyzed the verbalizations to, identify categories of terms that could reflect the formation of the, clusters. ASA-SIAM Series on Sta-. Bhatia, Department of Psychology, University of Pennsylvania, D22 Sol-omon Labs, 3720 Walnut Street, Philadelphia, PA 19104. Read the latest articles of Journal of Experimental Child Psychology at ScienceDirect.com, Elsevier’s leading platform of peer-reviewed scholarly literature Shakespeare, W. (1992). In the second step, the clusters extracted from these den-, drograms were interpreted by submitting the descriptions of the. indicated how well the participants identified the sounds. berger, Herrera, & Celma, 2004). We, report the detailed analyses of the primary dendrogram here. (2002). Auditory detection of hollowness. Context theory of classification, Misdariis, N., Minard, A., Susini, P., Lemaitre, G., McAdams, S., &, Parizet, E. (2010). (1997). These two types of organization (one based on a simple hierar-, chical structure and the other on a complex semantic network) are, either too simple to fit our perception of environmental sounds or, too complex and broad to be effective (e.g., in the field of post-, production, a sound designer should be able to efficiently search, sounds in sound databases). Therefore, our goal in this, article is to explore the categorization of environmental sounds. We used a two-step analysis to study and interpret these, classes. Considering the hier-. previous analysis, we calculated the occurrence of each term, in this case, the descriptions were tied to each cluster. Occ, Experiment 1. Following the sorting, task, in a second step, the participants were asked to type the prop-, erties shared by the sounds in each class. (Susini, Houix, Misdariis, Smith, & Langlois, 2009): • Object and action involved during the production of the, in the five semantic fields. Journal of Experimental Psychology: Learning, Memory, and Cognition Putting Bandits Into Context: How Function Learning Supports Decision Making Categorization is a cognitive process that unites different enti-, ties of an equivalent status. to this first residual matrix, producing a second residual matrix. Our statistical analyses were based, on a particular hierarchical cluster analysis producing two dendro-, grams (Hubert et al., 2006), allowing us to focus on more homo-. and the corresponding actions (leaking, fire). solids, but it remains to be tested for liquids and gases. The exemplar theory can be, viewed as a generalization of the prototype theory with multiple, However, perceptual similarity alone cannot explain category, membership because, for example, context can change the prop-, erties on which the similarity is based (Tversky, 1977). The role of similarity in the development of. Sound effect, taxonomy management in production environments. The progress of this research and the results of the database search engine built on this similarity model will be presented and discussed. Experiment 2. Environmental sound perception: Metadescription. In Order to Read Online or Download Journal Of Experimental Psychology Human Perception And Performance Full eBooks in PDF, EPUB, Tuebl and Mobi you need to create a Free account. temporal aspects of the sounds (patterns, etc.). (2007) also tried to explain the structure of the categories with a, multidimensional scaling representation of data from judgments of, similarity between pairs of sounds, but the interpretation of the, We used two specific techniques to overcome what we identi-, fied as potential limitations in existing studies of environmental, sound categorization. egies (see Appendix B for more detail). Grassi, M. (2005). Since this model is based on numerical results derived from experiments that mostly used synthesized sounds, the application to a database of recorded musical instrument sounds needs a strong validation procedure. The main clusters are indicated by Latin. Our structure of categories may be useful, for supervised classification based on acoustical descriptors, like, the work by Misdariis et al. of a mechanism, glass, and metal objects, and large mechanisms. Drioli et al. Many, subcategories were available for the subjects to categorize these, Experiment 2: Classification of Solid Sounds, The results of the first experiment showed a general level of. This cluster was close to Clusters 9, 10, and 11. Kim, N. G., Effken, J. Participants can, classify the same set of sounds according to either the objects or, To classify sounds, listeners must rely on pieces of information, that invariably specify these properties. Part of the evidence for this fascination can purpose of the study) who assigned a label to each category. Cluster 1 was distant from the other two large clusters (2, 3, 4, 5, 6, and 7, and 8, 9, 10, 11, 12, and 13), but the general analysis, of verbalizations showed that this cluster shared a lexical field with, Cluster 7. The analysis has four steps. The information that listeners focus on depends on both, the listener’s expertise and the identifiability of the sound. its properties when listening (Cummings et al., 2006; Lebrun et al.. 2001). The stimuli were, amplified dichotically over a pair of Yamaha MSP5 loudspeakers, (Yamaha Music Europe Gmbh, Rellingen, Germany). The second (and more spe-, cific) analysis of each cluster also resulted in the distinction. In sum, we found at this general level that the lexical fields, describing the main clusters reflected the main categories in Gaver’s, (1993b) taxonomy. Using such a procedure ensured that the primary den-, drogram was not contaminated by other marginal strategies. A hierarchical decreasing clas-, sification is applied to this new matrix through an itera-, of each CRV. Barsalou, L. W. (1983). Journal of Experimental Psychology: Animal Behavior Processes \ VOL. representation (e.g., the preparation of meal in a glass or container. However, this organization did not include gas and, liquid sounds. The Foley grail: The art of performing sound for film, Nearest-neighbor generic sound classification. E-mail: olivier.houix@ircam.fr, Journal of Experimental Psychology: Applied. Some of the differ-, ences between the results of the aforementioned studies could be, explained by the different emphasis on different aspects of the, Second, different mathematical representations were used. pants, except for numbers 31 and 32, which correspond to matrices generated randomly. Occ. In both cases, we aggregated the descriptions of the. The first lexical analysis, produced the hierarchical structure of the lexical fields related to, clusters. E-mail: bhatiasu@sas.upenn.edu This document is copyrighted by the American Psychological Association or one of its allied publishers. rization of the sounds. In Experiment 1, we examined the generative and abstract aspects of transformational invariants. A, close inspection of the second strategy showed two modalities of, categorization, which were interpreted as the result of two different, cognitive processes. (1997) and Dubois (2000): listeners identified action at a basic level and at a subordinate level, the source. When context, changes, different properties are engaged in the similarity be-, tween members of a category. This is also, true for the sounds caused by hand actions, such as clapping, or, mouth action-related sounds, such as laughing, nonaction-related sounds such as water boiling (Aglioti & Pazza-, The selection of the sounds for our experimental studies is a, crucial choice that influences their categorization, and this selec-, tion has been taken into account with the other methodological, This section discusses the methodological considerations that, may influence the results of the studies, such as the instructions, given to the participants and the choice of the statistical and lexical, analyses. The organization between the, , etc. We observed many subclusters for the, sounds of solid objects. The second experiment focused, only on the sounds made by solid objects. Journal of Experimental Psychology, 3, 1-14. The study reported here followed this article, and investigated in more detail how lay listeners organize sound. descriptions of the category are associated with this new sound, which then belongs to the taxonomy. to be a useful contribution to the analysis of classification tasks. (2000) developed a specific technique to analyze. be labeled as hybrid sound events (liquid and solid interactions). Our analysis of this apparatus is what this book is about. Human sensitivity to acoustic, Cano, P., Koppenberger, M., Herrera, P., & Celma, O. Clusters 5 and 6 distinguished rotary mechanisms such as gears, (Cluster 6) and rotary mechanisms with objects returning to their, initial position (Cluster 5). categories of sound sources: solids, gases, liquids, and machines. Portraits of the Subclusters. The sound levels were ecologically adjusted to reproduce the. chi-square indicated that the terms were specific to this CRV. B. Lloyd, Ross, B. H., & Murphy, G. L. (1999). This result is, similar to the distinction between liquids, solids, and gases pro-, At a finer level, the subclusters tended to correspond to the, different types of actions that produced sounds. The listener must, isolate sound events from other concurrent events and integrate the, different sensory inputs. the sounds in a number of classes that varied between six and 28). The 60, sounds were monophonic recordings of events usually occurring in, a kitchen. Rips, L. J., & Collins, A. The different operational processing, stages (from sound encoding to attribution of meaning) not only, form a bottom-up process but also incorporate previous experi-, ences of the auditory world and information from the other sensory, modalities in a top-down process (see also the general model, proposed by McAdams, 1993). Nevertheless, we observed a, large distinction between four main clusters of sounds: the sounds, produced by solids, liquids, gases, and machines.