1.1. Place/space

A variety of meanings have been attributed to the word ‘space’, as it appears in the literature of diverse disciplines and research fields. In considering the specialised interests of this investigation, space can be regarded as a generalised set of dimensions, encompassing objects of relative size and distance, that does by no means imply that there exists any other kind of relationship between these objects; it is merely a reference framework, or ‘the dimension of quantitative divisibility’ (Massey Reference Massey2005: 23). Should a spatially related set of objects appear to be linked by action, however, it is attributed with significance; its constituting objects appear to suggest utility, or any other property that relates to some meaning. A bedroom’s bed, pillows, sheets and curtains are related through the action of sleeping, for example, and thus the bedroom space becomes place, as it has acquired a purpose. Tuan argues that ‘what begins as undifferentiated space becomes place as we […] endow it with value’ (Tuan Reference Tuan1977: 6). It is possible to generalise and assert that the quality of place is linked to activity and, as such, also encompasses time. In the words of the leading figure of the Situationist International, Guy Debord, places are made by ‘subjecting space to a directly experienced time’ (Debord Reference Debord1995: 126 – thesis 178).

The sense of place, consequently, emerges as a result of a more complicated interaction than mere observation: It is the consideration and appreciation of perceptual stimuli that provide this general impression, through association and other cognitive processes. As will be discussed later, the aesthetics and poetics of visual stimuli, such as the observation of architecture and landscape, can be important; so, of course, can sound. A method of representation of the sense of place using sound as the sole medium must, therefore, provide means of reproducing general sensations and impressions, complete with context, memory and emotion involved as appropriate.

1.2. Psychogeography

In his ‘Introduction to a Critique of Urban Geography’, Debord defines psychogeography as ‘the study of the precise laws and specific effects of the geographical environment, consciously organised or not, on the emotions and behaviour of individuals’ (Debord Reference Debord1955). The inverse, obviously, also holds: Situationism ‘presupposed that it was possible for people to synthesise or manage these situations [which affect the individual’s consciousness and will] as an act of self-empowerment’ (Sadler Reference Sadler1998: 45–6). The relationship between individuals and environment is therefore bi-directional. Psychogeography is thus the interaction between human and space and, as such, also a study of place.

The situationists built maps (dubbed psychogeographical) that ‘gave all power to subjectivity, […] expressed insubordination and chance rather than certainty’ (Merrifield Reference Merrifield2005: 48), such as Debord and Jorn’s 1957 Naked City (Sadler Reference Sadler1998: 60). The maps were in fact collages of pieces of conventional maps interconnected with arrows that suggested flow between, and affinity among, places, ‘odours and tonalities of the cityscape, its unconscious rhythms and conscious melodies’ (Merrifield Reference Merrifield2005: 31), ‘[investigating] the relationship between language, narrative, cognition’ (Sadler Reference Sadler1998: 60). The endless possibilities of navigation presented by the arrow networks reflect the activity of dérive, or drift. The latter, ‘a mode of experimental behaviour linked to the condition of urban society’ (Merrifield Reference Merrifield2005: 30), was the process of walking around cities guided by psychogeographical criteria and documenting human activity and its impact upon places. The maps would often give way to subjectivity to a surreal extent, as was the case with Constant’s cityscapes, ‘redolent passageways, shocking landscapes, superimposing routes and spaces onto each other’ (Merrifield Reference Merrifield2005: 51). In these cases, the maps would function as representations of imaginary landscapes,Footnote 1 visions with no immediate equivalent to actual geography. These maps, due to their close relationship with the dérive, may provide a promising representational model for this project.

1.3. Impression/expression

The relationship between sound and vision is strong but delicate. One hints at the other in constant fashion, contributing to our environmental understanding. It is not uncommon for the two to not concur in a singular impression. An obvious cause for this disjunction of the senses is sound’s ubiquitous presence: sounds cannot be ignored directionally in the way that sight can be directed in space. Furthermore, a dream, or an expression, of a site might have a variety of sounds associated with it: evoked from materials, apparent spatial features such as distance and size, even sonority from objects that appear capable of producing sound. Inversely, a sonic stimulus can evoke the impression of an associated visual image.

An image (not necessarily visual) that possesses such evocative qualities can be said to be poetic, in the Aristotelian sense. Bachelard remarks that:

Therefore, an object’s image is a being in its own right, a separate entity from the object’s material existence, and disjunct from all causality related to the nature of the phenomenon. It is possible to consider this image (or general reconstruction, in the sense discussed) as completely remote from the reality from which it originates. Bachelard argues that an interpretation of such an image should not be attempted, as doing so will cause it to be ‘[translated] into a language that is different from the poetic logos’ (Bachelard Reference Bachelard1964: xxiv).

The experience of imagery can thus be broken down into two stages:

1. (i) Impression: the reception of the stimulus and the subjective reconstruction of a poetic image, complete with any mental transformations as previously discussed. The image is appreciated and a primary sensation is established, such as affect, allure, resentment, intrigue or otherwise.

2. (ii) Expression: The image is formed into an experience that can be described and represented.

It is our hypothesis that evocation takes place in the very early impression stage, followed by the effects of consciousness, while context and association occur later to form the expressive image.

1.4. The psychosonographic

What is hereby attempted is a sonic representation of place as an expression of a mental image, which is a result of sensory experience and is causally unrelated to spatial materiality. Psychogeography, its maps and the dérive lend themselves well to this end. The resulting representation will therefore be a product of drift, the determinants of which are:

It can be asserted that the general form of this project is the sonic equivalent of a psychogeographical map. What is sought are the ‘fissures’ between places that are uncovered through the process of impression and expression, disregarding any geographical proximity. Psychosonography is introduced to encapsulate this process, while a realisation of this type of sonic map (i.e., a psychosonographic representation) will henceforward be referred to as a portrait.

Consequently, there exist two varieties of psychosonographic drift: the actual, equivalent to the situationist practice, during which recordings may take place, and the virtual, which is a result of using a psychosonographical map, a purely mental process. The drifters are different persons: the recordist and the wanderer, the author and the character of a narrative, the designer and the user of a map.

The use of location recordings may suggest a similarity to soundscape composition. The latter, however, deems ‘the acoustic ecology arena as the basis from which [it] emerges’ (Westerkamp Reference Westerkamp2002: 52), and, as such, serves a completely different purpose. It is reasonable to suggest that soundscape composition is to psychosonography what photography is to expressionism: the Brücke artists, for example, ‘tried to capture their sensory experiences and visual impressions […] in the form of paintings’ (Elger Reference Elger2007: 12). Also, Wassily Kandinsky, co-editor of the original Blaue Reiter publication, in his ‘Concerning the Spiritual in Art’, wrote of ‘artistic recreation of mood in its inner value’ (Kandinsky Reference Kandinsky1914 – also cited in Kahn Reference Kahn1999: 107). Psychosonography and expressionism share an attitude toward representation of material reality: impressions and experiences are rendered directly, with a referential and not interpretative perspective on the actual nature of the cosmos, with a focus on the subject of observation; they make use of the ‘poetic logos’.

2.2. Mimesis and imagery

In order to use the portion of this vocabulary that does not originate from the location under description, as explained above, it is necessary to discuss the particular aspects of the intended mimetic discourse. Emmerson uses the term ‘mimesis’ to ‘denote the imitation not only of nature but also of aspects of human culture not usually associated directly with musical material’ (Emmerson Reference Emmerson1986: 17), a particularly applicable definition for our purposes. He also specifies two types of mimesis: ‘timbral’ and ‘syntactic’ (Emmerson Reference Emmerson1986: 17–18). The simplest form of mimesis can use both at the same time, an instance frequently encountered in instrumental music: an instrument attempts a direct imitation of a natural feature (a flute imitating birdsong, for example, relies on both timbral similarity and its ability to approximate gestural content). This is another reason for the choice of instruments, in that they offer vast potential of gestural control. Smalley considers instrumental gestures as ‘stand-ins for non-musical gestures’ in ‘first order surrogacy’ (Smalley Reference Smalley1986: 82).

It is important to clarify that this is a particular use of the term mimesis, which is also traditionally associated with resemblance (as noted in Halliwell Reference Halliwell2002: 15): it does not necessarily imply the process of imitating but the relationship of the represented to its original. Plato is concerned with the representation of essence (oυσί1;α): In his Republic, he argues that the existence of an object is twofold: it exists materially and as a concept (ιδέαFootnote 3 – Plato Reference Plato1992: 104, p. 596b) that represents philosophical truth and is, in effect, an archetype. A craftsman can create an object by considering this concept, but cannot craft the concept itself. A painter, in a similar fashion, can also create the same object through representation, but cannot be credited as the object’s creator (Plato Reference Plato1992: 597e–113). In short, Plato suggests that a representation is an imitation of matter (appearance – ϕαινóμενoν), which is an imitation of truth (being – oν) (Plato Reference Plato1992: 598b–113), and therefore is removed from essence – an imitation of the second order. Plato’s argument also poses a requirement for the representor to possess some knowledgeFootnote 4 about the particular representation, and not merely mimetic skill. The process of the dérive addresses this condition in part. It becomes necessary to consider the nature of the imagery we are hereby attempting to create.

We are enforcing a restriction to the senses: we only represent place with sound. We have established that a direct, phonographic mimetic approach of the first order does not suffice; its reality cannot be in accordance with the essence of its source (the philosophical truth). Thus, we are in need of an additional layer of discourse, in order not to represent, but to reconstruct the experienced reality that we seek to portray with respect to the aforementioned essence, a necessity that we have already catered for by considering the use of musical instruments, as well as other arbitrary sounds. By presenting a variety of cues for the missing stimuli (related to senses other than hearing, or even sounds that can not be captured or do not exist) and characteristics of place that are not directly linked to human senses but are results of experience, this new reality does not seek to contrast the original one, but to blur the distinction between them (within the space-time of the artistic product) by approaching the archetypal, the general form, the philosophical truth. The latter is approached through abstraction: By transforming (in time or in stasis) a phonographic image into one that does not fit the original context (in Wishart’s terms, creating an imaginary landscape of the type real-objects/unreal-space (Wishart Reference Wishart1996: 146) – in a similar way to psychogeographical maps as discussed in 1.2), we are predisposing the listener to perceive their common attributes, thereby asserting that we portray an aspect of place that does not depend upon its materiality. The effect of combining (sequentially, simultaneously, or in a temporally remote manner through memory) this expression with evidence of material reality creates an ambiguous state, a model of some heightened perception that can occur in what Baudrillard dubs the hyperreal, ‘sheltered […] from any distinction between the real and the imaginary’ (Baudrillard Reference Baudrillard1994: 3). Mimesis, in this context, becomes simulation, a process of imitation that ‘threatens the difference between the “true” and the “false” ’ (Baudrillard Reference Baudrillard1994: 2).

Furthermore, it is stressed that, in applying the above rationale, the subject of expression is the dérive itself, not its records (in the form of audio recordings or otherwise). Were it not so, it would appear that such a representation would be a Platonic imitation of the third order, even further away from philosophical truth. However, in practising a process of selection of attributes of place by experiential criteria and in portraying them, stripping ornamental references to their materiality and context, we are approaching their original formFootnote 5 and thus arriving at an image, with severed causal bonds to its origin, that is a being on its own and, according to our previous statements, is of a poetic nature. The analogies among these levels of mimesis are reflected in figure 1. Although the imaginary originates from an imitation of reality, it is closer to an archetypal form through abstraction. The side view is the perspective of the hyperreal: reality and imagination are indistinguishable.

Figure 1

Mimetic relationships.

2.3. Composition

Since it is required that a sound exhibits a certain attribute that relates to an object’s formal origins (its essence), it follows that it must either be processed in order to acquire this attribute or possess it to begin with. Most DSP-related transformation processes, however, tend to distort the sound’s landscape – its ‘imagined source’ (Wishart Reference Wishart1986: 44) – towards the unreal. It will be difficult to establish mimetic discourse, as discussed, in this way, because this attribute will appear to originate from an unidentifiable source and will not lend itself to the image we are attempting to establish; it will be, in this respect, alien. On the other hand, in the event that a sound possesses this attribute by virtue of timbral or gestural characteristic, it needs no major transformational process and can be used directly. In this case, the sound retains its original landscape, which, in the case of an offsite sound, will be different to the landscape of the portrayed place by definition and, in conjunction with a reality reference, an imaginary landscape of the type real-objects/unreal-space is established, analogous to the psychogeographical maps as argued in 1.2. In this case, as there appears to be no particular reason for the object’s presence in the portrayed image, mimetic discourse is arguably easier to be perceived. As an example, we can consider Wishart’s transformation of the word ‘Liss-’ (from ‘listen’) to birdsong from his Red Bird (Wishart Reference Wishart1996: 156–7), which is clearly only possible because of the timbral similarity between sibilance and the formants of the voices of the birds. That is, the sounds share a morphological thematic link (Parry Reference Parry2000: 45); the imaginary landscape of the combined sounds is that of a voice imitating birdsong with an imaginary degree of accuracy. The temporally inverse transformation would suggest a bird that can imitate talking.

In a similar manner, we may accomplish a rudimentary transformation by masking the sound with the actual soundscape and gradually aiding the recognition of its original source. This provides a time-domain compositional axis: source recognition. It becomes straightforward to assert that, in such a continuum, there exist ambiguous states where the initial reference frame appears altered but the cause of this alteration is not apparent. In this case, when the source reveals itself, the ambiguous state is resolved, in a manner similar to a cadence. The precise point of identification can be seen as a pivot, around which the particulars of our mimetic discourse revolve. Parry proposes the term ‘mimetic value’ to describe the ‘position of a sound on the axis of source recognition’, derived from Ten Hoopen’s ‘specific/surrogate’ continuum (Parry Reference Parry2000: 49–50), a proposition we shall adopt. Figure 2 depicts this relationship between the mimetic value of an offsite sound object and the ambiguity that results by juxtaposition with location sound, resulting in the definition of a composition space (the enclosed area plus the linear section to the right, where ambiguity is resolved). It must be stressed that ambiguity refers to the resulting state, rather than the sound in question itself. Likewise, mimetic value is a property that depends on the morphological relationship between the sound it refers to and any other sounds involved in the particular state. The pivot can be seen as a point on the recognition axis, relative to this relationship, and can occur at any time. More importantly, it needs to be crossed from left to right – mimetic value needs to be increasing. Otherwise, memory prohibits ambiguity from ever occurring, as the object will have been already identified. Mimetic value, once it increases past the pivot of recognition, cannot decrease again under normal circumstances. Such ‘reverse’ movement can, however, be employed, as the thematic links will be intact.

Figure 2

Mimetic value and resulting state ambiguity.

There exists a special case: if the pivot of identification happens to indicate a very low mimetic value, it is conceivable that a sound might cross it during its very early attack phase – this is particularly applicable to musical instruments that, masking effects aside, are intrinsically recognisable and, as such, exhibit a normally high mimetic value. In such cases, ambiguity occurs only during those initial milliseconds of the sound’s life. In 1.3, we stated that evocation takes place in the very early impression stage that we associated with first reception of stimuli and what Bachelard refers to as onset, and plays an important part in the establishment of the poetic image. Assuming that our expression is perceived by a listener, who is therefore in impression, this miniscule attack phase is enough to establish an image. This is the reason that certain sounds, although clearly identifiable as alien to the portrayed location, sound as they are somehow a part of it: it is still possible for the onset of the sound to have a thematic link to an object that either is present in the environment or has some potential to be so. For example, although an overblown pan flute (to the extent that it produces harmonics) can under no circumstances be mistaken for a train whistle, they share the characteristic of the overblown pipe and, as a result, the abstract quality of the steam engine – ‘train-ness’ – can be consciously or otherwise evoked. Another way to regard this relationship would be as a thematic link of the second order: the flute shares a morphologic link with a train whistle, albeit only at the very early stages (which is apparently long enough for the link to form), and the train whistle, an intermediate stage which is never heard in the process, shares a mimetic linkFootnote 6 with a train’s steam engine. Arguably, the resulting signFootnote 7 ‘pan flute/steam engine’ is more suitable than an actual (modern) train signal in conveying a sense of place. The latter is not reminiscent of a train at all: it is designed to signify a general warning. In other words, such sonic signs can signify any concept we wish to express that relates with our subjective experience of place, as long as their signifiers can operate within our proposed framework. This is one of the mimetic schemes used for Thessaloniki’s railway station in Urban Portrait.

As another example, in our portraiture of Thessaloniki’s seaside, we have used flute sounds to imitate wind, while exploiting their sonic affinity with clarinet multiphonics that are meant to embody the poetic representation of distant ships. The association ‘clarinet/ship image’ is formed by a subjective interpretation of general impressions such as weight on water, rust, and slow motion. What is conveyed is not the abstract notion of the ship, but the expression of a personal experience of one when viewed from a certain place and under certain conditions of observation. Strictly speaking, this is not a sign, as there exists no definitive relationship between signifier and signified. Once, however, we remove the objective notion of the ‘ship’ and consider the immaterial sensation that results from its sighting, we may speak of a sign of the general form ‘signifier/poetic image’. While this proposition does not make linguistic sense, it allows for a subjective resolution of the relation between the sonic expression and the image, thus enabling abstraction from immaterial stimuli.

2.4. Place sounding

In attempting to solidify a method of sonification, we assert that mimesis, as discussed in 2.2 and applied in 2.3, is our primary tool. The subject of our expression is our impression of the places we are attempting to express, as established during the drifting process and experience. The object of expression is another impression, one coming from an imaginary dérive, undertaken by a fictional character that we expect our listener to identify with: our expression resembles storytelling, other than the fact that we are concerned with imagery rather than narration. To this end, we can use mimesis to highlight our experience of place, using a record of it (in soundscape recordings) that can be enhanced, effectively populating the sonic expression with: (i) objects that exist in space, the impression of the existence of which we wish to highlight by elevating the perceptual importance of the features that approach the objects’ form; (ii) objects that exist in space but produce virtually no sound, or objects that do not exist at all, by presenting a surrogate structure that functions in a similar way; (iii) sonic signs that associate with a feature of place or its poetic impression that cannot be captured by recording, due to a silent nature or immateriality; and (iv) abstract sonority related to mood and ambience, by employing sonic forms that we believe to evoke a poetic image similar to the one we wish to express.

3.1. Drifting and recording

As thus far elaborated, the process of recording is the beginning of this endeavour. What is perhaps the most significant issue, other than a problematic acoustic ecology associated with urban environments (‘lo-fi’, to use Schafer’s terminology (Schafer Reference Schafer1994: 43)), is that it alters human-related features of place. Like any kind of observation, it disturbs the phenomenon: people are reacting in one way or another to the sight of a microphone. The primary concern was not the sonic quality of the result but the transparency of the process, in order to ensure that what is being recorded is unaffected by the presence of the recordist. Therefore, most of the field recordings were made using equipment as discreet as possible, such as pocket flash recorders and miniature microphones, effectively sacrificing definition for fidelity (see (Chion Reference Chion1994: 98) for a discussion of the terms).

The drift itself, however, is of a more general import. As detailed previously, it is in this process that all sensory information is gathered in impression. Since we have embraced subjectivity as an expressive quality, drifting, during which all impressions are formed, is a part of the project in itself. It would be rather difficult for someone with no intimate knowledge of the city to construct this portrait. Thus, we can assert that experiencing the city in psychosonographic terms is central to the development of the project.

3.2. Form

This incarnation of Urban Portrait is about Thessaloniki, Greece. A total number of six places in the city that exhibit a particular interest were selected: the docks, the railway station, the old Modiano market, the seaside, the Upper City and the central Tsimiski avenue. After an initial assembly, selection and editing of location recordings, a number of local musicians were invited to a recording studio and asked to improvise on relevant environmental features, both before and after hearing the recordings. Despite a general guidance and a few prescribed phrases, they were free to produce any sound they felt appropriate, as long as they did not attempt to interpret the feature in question. Therefore, the combination of location and instrumental sound contains traces of the musicians’ impression and expression of place, in effect enriching the work and protecting it from accidental bias. The result of this process was an extensive compositional palette: bassoons and bass clarinets were used against boat horns and traffic sounds; the prepared piano proved an effective source of mechanical sonority; flutes and clarinets using multiphonic techniques provided an assortment of fricative or flowing textures related to various material elements and natural phenomena; and various recorded objects, such as kitchen utensils and furniture, contributed to the modelling of environmental aspects that are impossible to capture with location recording, to mention but a few examples.

In order to express ‘flow and affinity’ among the portrayed places (see 1.2), a set of transitions were defined. The latter connect the senses of place through elements of their expressive forms, without necessarily demanding any kind of geographical equivalence or proximity, as we have already stated in introducing psychosonography (1.4). Transitions are not an ornamental feature of this project: place can be important not so much because of its existence, but because of its relationship with other places. To facilitate a transitioning scheme, each place is fragmented into scenes that revolve around a particular theme. For example, the general location of the city’s docks is divided into ‘far view’, ‘ships’ and ‘machinery’. Different aspects of experience of place are divided into different scenes, each of which presents two possible destinations: one related to another scene of the same place, which is the default, and a second, optional transition that links the present place to another through some expressive similarity of their respective scenes. Since we wish for the listener to be able to interact with the piece by enabling him or her to instruct a transition, the adoption of this scheme effectively circumvents the problem of defining transitions with a many-to-many relationship by reducing it to a set of simpler problems.

It becomes difficult to describe the individual renditions of places in a general way, as each presents different challenges and necessitates a unique adaptation of method. As an example, a scene of the empty market building is populated by either salesmen or travellers, in which case it is transformed into the railway terminal. In another instance, the folk music inside a traditional Upper City tavern is used to transition to the street, through the recording of a busker and a musician imitating his style. Ambiguous states are also exploited, as they can be resolved into abstract space that can then be gradually contextualised. Such an example would be the sound of a train stopping on the platform, which resolves a previous complicated interplay among sounds of train tracks, a snare drum roll and clarinet pad taps. The latter then, through layering and digital processing, appear to densify and eventually transform to rain, thus preparing the ground for a stormy seaside scene. In the case of transitioning between two scenes that belong to the same place, the task becomes significantly simpler, as there exists no need for an intermediate state.

3.3. Installation

The installation is based on a computer running Max/MSP (Cycling ’74/IRCAM). The employed quadrophonic diffusion system provides ample space for the coexistence of the multitude of sounds that must be heard in simultaneity, greatly diminishing the need for additional filtering and significantly improving the usable dynamic range. The human/machine interface consists of a button that can alert the listener to the possibility of a transition by blinking a light-emitting diode on its base, with a subsequent press executing the transition. A display provides information on the place currently being diffused; it is placed in the centre of the room, along with the button.

In outlining the project’s form, we have attributed two possible transitions to a scene, dubbed ‘default’ and ‘optional’, with the apparent intention to provide the listener with a choice. This interaction materialises in two steps: the listener is informed that there exists an optional destination, which is obviously related to a sonic element appearing at about that moment that carries a hint about what the destination might be; then, for a pre-determined amount of time, the user has the option to instruct that transition, or eventually progress to the next default scene. This is loosely modelled after the dérive practice, where navigation ultimately depends on the drifter’s perception of environmental features. However, the importance of the – limited – interactive aspect of the work can be overestimated by the user, as a result of the interface presented to him: it is a common expectation for a press of a button to bring about some perceptible change. In contrast, the installation’s button is inactive most of the time and, even when it is not, the change it incurs is gradual and often begins several seconds into the future. This is a condition that limits the comprehensibility of the artwork and prescribes the use of operating instructions. To an extent, it also conceals the fact that this is primarily a listening piece and not interactive art as such. To counteract the problem, visual elements have been kept at a minimum, in an attempt to direct attention to sound as much as possible. During the piece’s presentation, it became apparent that providing seating would also encourage listeners to stay longer (the latter is important as some of the scenes are several minutes long). Furthermore, the use of statistical data of users’ transition choices was carefully considered. Indeed, they are processed and stored in a transition probability table, but utilisation of the latter was ruled out because, as with any statistical system, meaning only emerges given a sufficient volume of data, which in turn would necessitate a large-scale, long-term exhibition of the piece.