Margulis (2005) adds more nuance to the idea that unexpected events arouse tension: she is particularly interested in the types of tension that are aroused. Margulis has developed the model to distinguish the three types of tension: surprise, denial, and expectation itself. This model has informed the analytical work in the current research. It has proved fruitful in examining the differences in working of different pedals/drones. 
The predictions about the following events are based on what the listener (consciously or unconsciously) considers ‘the norms’, or ‘logical music continuations’ in any given moment. Everything else (the material that surprises us) is felt as deviation from the norm (or from what was expected). In Meyer’s writing deviation is an important concept. In the context of this thesis, it could be associated with the concept unrealized expectation.
Being acquainted with the musical style does not only influence the listener’s expectations on the note-to-note level. Each style has its own characteristic patterns that could be recognized as such. A composer can refer to another style by ‘importing’ these patterns in own composition. Take, for example, Beethoven’s Pastoral Symphony. The listener can recognize its folk drones as such, if she is acquainted with the concept ‘folk drones’. The meaning of open-fifth sustained tones in this piece is created through recognition of this model, thus, through understanding it as a particular pattern – in this case originating from a different musical style, with different stylistic norms.
The importance of interrelations between patterns and the knowledge of norms is recognized also by James Hepokoski and Warren Darcy, in their influential work Elements of Sonata Theory (2006). They consider each piece in sonata allegro form as being in dialogue with all the other already existing pieces in the same form. The word “dialogue” is very interesting here. It suggests the interaction between the general and the particular. The general influences the perception of particular; the particular contributes to the ever-changing identity of general. Hepokoski and Darcy define five types of sonata form, which are described as models. They also define an ordered number of ‘default levels’ for tonal plan (and types of cadences), which could also be understood as models. In the authors’ view the composer is in dialogue with these existing models. In the analytical chapter ‘Models’ of the current thesis I will argue that there is a number of pedal and drone models, with which any particular instance of pedal or drone is similarly in dialogue. When any particular pedal/drone model is recognized in a piece, the ‘dialogue’ that relates it to the actual pedal or drone in that piece influences our perception of that tone. Each model is in fact a stylistic cliché. Even when deviations (unexpected or postponed events) are defining a particular type of pedal on a note-to-note level (e.g. in case of a dominant pedal), on a higher level, a type of pedal could be perceived as a part of the norm.
Some types appear on particular places within formal structure. The listener is sometimes able to orientate in the form when hearing some characteristic realization of pedal. For example, in a Baroque fugue, hearing a long dominant pedal probably means that the composition is in its final stage; in a classical sonata form, such a pedal would probably make us expect the recapitulation. The combination (or interaction) of the pedal tone and the formal structure influences our perception of the piece. Whenever a pedal makes us expect something, it implies that the meaning of the other (simultaneous) notes is changed. The meaning of both pedal and other notes is influenced by their interaction.
All the tones in a piece, with their properties, are integrated in a complex network where the levels of meaning are interwoven. Interaction between diverse types of musical events (including pitch, dynamics, clichés, etc.) influences the way we perceive musical patterns, and is also the ground for the forming of meaning. Through interaction with other musical elements, a sustained tone could be understood as being a particular type of drone or pedal (or none of them). In the chapter ‘Models’ I will outline and discuss a number of pedal/drone types (models). Their ‘working’, or the power that each of them can generate, will be seen as related to three things: theory of expectation, physical properties of the sound, and interaction between the musical parameters.
To a certain extent, the ‘working’ of pedals and drones is cross-stylistic. For example, in any style or genre with tonal-functional harmonic language, a dominant pedal will probably raise tension. Also, a drone with powerful, low sound will, in any style, probably cause some thrill. But, there are also many style-dependent effects. For example, a single interval can be consonant (stable) in one style, and in another style, we would expect to hear its resolution (it would raise tension). Style-dependent expectations could also relate to form, to harmonic progressions, to any kind of common patterns. In pop music, for example, a short recurring pattern, riff is very important. This pattern (that could be based on a pedal tone) is, in the listener’s musical mind, instantly being associated with the song it comes from. The repetition of such pattern, in the context of a pop song, could then not have the same effect as a temporary pedal-ostinato pattern in a western art-music piece. The meaning we assign to musical gestures are always related to stylistic norms that we have consciously or unconsciously taken as the frame for the listening.
These considerations have proved to be very important in defining pedal/drone models. In order to narrow otherwise too broad field of research, I have primarily focused on the classical repertoire - which makes the core of my analytical chapters. Nevertheless, I have not restricted myself to it. The classical repertoire is already very rich in variety of pedals and drones, but at certain places it is necessary to refer to styles that are related to a particular type of pedal (such as folk drones or drones in ambient music). On other places, I have included examples from other styles with intention to emphasize the sameness, or to suggest the difference where similarity seems to be high. My ambition here is only to sketch the possibilities for further exploration. More research would be needed to examine the specific working of pedals and drones in the context of other musical styles.
The methodology of my research includes three lines: 1) the line of the literature, especially that related to the expectancy theories, 2) the line of listening to broad repertoire of music, and 3) the line of score analysis. The three lines were interwoven throughout the research: each of them has been influencing the others. For example, the ideas about music expectancy have modeled my analytical work; performing auditive research (listening while focusing on different musical events, comparative listening of various musical fragments) has influenced the definition of each model; the conclusions made on the basis of listening and analysis have occasionally led to changing of the focus while re-examining the expectancy theories; different theoretical focus has provided different language for the analysis.
After this introduction, the first chapter provides a brief overview of literature related to the topic. In the second chapter, I will focus to the concepts of pedal (pedal ostinato) and drone, seen from the perspective of norm-deviation idea. I will outline and explain the concepts related to expectancy, that will be used in analytical chapters. The third chapter deals with the concept interaction, and its possible role in analyzing appearances of pedal/drone. The central part of the research is presented in the chapter on pedal/drone models. Each proposed model will be defined and illustrated with one or more analyzed musical examples. Many instances of pedals and drones are not clear examples of one or another model. Therefore, I have also included links to a number of short ‘pedal/drone stories’, analytical presentations of various interesting examples. At the end, there are three case studies, where the chosen pieces are explored into more details.
At last, the electronic medium made possible to include the sound files, and larger portions of the full scores. Therefor I have chosen not to make the reductions but instead to annotate the scores. Clicking on the small image within the text will open the score in the full format.
Next: Pedal And Drone In Literature: A Very Short Overview
 Expectancy theories can help explain much of the tension the pedals create, but musical expectation is certainly not the only factor that contributes to their power. The sonorous quality of the sustained tone can have tremendous effect. One of the critiques that Mayer receives is that he “underestimates the connotative or denotative aspects of sound. Individual sounds can sound "cute," or "crude," "nervous," or "assertive," and these connotations influence the emotional experiences of listeners.” (Huron, Course Notes for the subject Music and Emotion, chapter on Leonard Meyer, online). Spitzer is on the same line with Huron: “Meyer never extensively entertains the possibility that emotion may be produced through states rather than processes; or through the appreciation of regularities (such as ‘grooves’) instead of subversions.” (Spitzer, 2009).
 The unexpected (or deviating) elements will in most cases be re-interpreted to fit a new understanding of the music. In retrospect, most of the events would not be found illogical.
 With pedal-ostinato I refer to a rhythmic pattern realized on a sustained pitch. It could also be ornamented with other pitches.