The Trace of the City’s Structure: The Evocative Representation of Settlement Through the Montage Technique

Abstract

In architecture, it is now a consolidated practice to involve a process of critical approach and historical knowledge before the design phase. This need is even more evident in reconstruction contexts, such as the Old City of Mosul, where the project is required to return some of the city's identity features. That practice, conventionally understood as the passive collection of data, is however susceptible to a more active approach, the production and interpretation of a conceptual and abstract condition of the city, capable of also capturing the data not immediately visible and latent. Mosul presents an urban structure composed essentially of primary monumental elements around which spontaneous residential fabrics have developed, almost disordered in their overall appearance as a solid mass, only excavated by open spaces. The proposed experiment concerns the creation of a cognitive map, a virtual state that recognizes and enhances the generative settlement structure of the city starting from certain type-morphological invariants—the introverted character of the enclosed settlement; the structure of compensatory open spaces such as courtyards and patios; the density of the settlement with a prevalent horizontal development; the scalarity and hierarchy between public and private spaces, etc.…—that identifies and represents the modalities of an otherwise indeterminate spontaneity. The result is the reconstruction, or reinvention, of the meaning of places through a methodological instrument of investigation suspended between the recognition and interpretative condition of a survey and the limited structure of a possible design conformation. If the reconstruction cannot be satisfied with mimesis but requires an evolutionary and transformative vision of the city, this tool is configured as the link between identity and transformation, a synthetic act between knowledge and the inductive capacity of the project to emphasize some of the authenticity assumptions of memory, even in an evocative and conceptual reworking.

1 Structure of the Paper

The paper reflects on the theme, vital to urban reconstruction, of grasping and representing the memory and identity of places. To explore this topic, several historical maps were selected, along history, for their ability to deform geometric data and highlight some particular features: even if unrealistic, the value of those maps is, indeed, their capability to select and abstract particular elements. Since the theme of this contribution is urban memory and identity, the study of the applicative case (the Old City of Mosul) implied the selection of some characters, the invariants, whose role is to transmit immaterial heritage in new design practices. In the chapter about methodology, the procedure of selection is carried out with examples and leads to the creation of a virtual map of the city that, albeit unreal, aims to represent exactly those invariants mentioned. The paper proceeds to identify possible ways in which this method of knowledge can be applied in design practices through some case studies and then reflects on the applicative case of Mosul.

2 Introduction

The sensitivity—and consequently the attention to a methodological recognisability—of the design of the new in a pre-existing context (ancient or otherwise) is a more recent tendency than it seems, given the now common acceptance of this postulate; in addition, it represents an operational phase of the same preliminary investigation phase when one is called upon to intervene in an existing context: the analysis, which is already a genuine design activity. However, historically one can recognize other aptitudes: if “before modernism it was believed that cities had evolved over time, according to a process of imitation of natural laws”, and therefore did not foresee a particular protocol in their modification as, precisely, organisms in natural growth, “for the architects of the beginning of the twentieth century, the possibility of a clinical intervention in the historical and natural evolution of the city was out of the question. […] And they went to attack the bastion of evil, identified in the nineteenth century city” (Eisenman, 2014, p. 223). What Eisenman recognizes in these passages is already rooted in Vidler’s reading (Delevoy et al., 1978), when a genealogy of typologies is proposed: a first typology, of natural inspiration (meaning the Enlightenment vision, of architectural models derived from nature—trees evolving into columns, etc.); a second functional typology, based on the intended use of the buildings (the extreme that Eisenman defines as aseptic); and a third typology, of which Vidler considers himself an ambassador, autonomous and based on architecture itself, which initiated typological studies as they are understood today. In the essay, Eisenman retraces and interprets L’architettura della città by Aldo Rossi (1966)—universally considered a pivotal figure for this type of study—and identifies him as an experimenter of a new approach to the city, poised between a humanist-style subjectivity and a modernist objectivity. This is due to several aspects of Rossi’s research, first and foremost the type, a new entity of mediation between architect and project, a device in which resides, according to the proposed reading, both a dose of potential transformation and invention, and the guarantee of belonging to the architectural and building tradition; a second key point, however, is the introduction of a semantic difference between history and memory: “History exists as long as a subject is in use; that is, as long as a form remains in relation to its original function. But when form and function are split, and only the form remains viable, history passes into the domain of memory. Where history ends, memory begins. […] One can then say that urban history is the process by which the city has received its imprint of form, while the succession of events constitutes its memory” (Eisenman, 2014, p. 229). It is precisely memory the device that allows a clinical intervention in the historical city, which guarantees the lawfulness of such an operation compatible with history without running the risk of historicism and falling into a pathological inactivity of form.

3 Objectives

The introduction of an abstract character, such as memory—that is essentially unquantifiable—however, introduces the question of how this historical and urban memory can be represented by means compatible with the discipline of architecture, and thus not only through a text, but through drawing. Cartography, especially since the Renaissance, has increasingly developed the aspects of exactness in the representation of reality (analogous to the technique of perspective drawing), but this exactness seems to be insufficient for representing a further substratum of meaning, stopping instead at the inanimate skeleton of the urban structure and its condition of tangibility. In cases of urban destruction due to wartime conflicts, it is in itself evident how the representation of reality is only one of the tools useful for reconstruction; however, lacking information about meaning and memory, one would arrive at the paradox of seeing in the destructions only empty, equipotential areas, substantially deprived of the value of the locus of which they are instead bearers. If we can take for granted the preservation of certain permanencies, such as the street layout, it is not necessarily the case that this information is exhaustive of the argument of the intangibility of memory. What we wish to propose here is the identification of a methodological approach that seeks to represent—as abstractly as conceptually, albeit geometrically plausible—certain aspects of memory. All this is in the conviction that, given the objectives, there is not necessarily a contradiction of terms between authenticity and reality, or rather: between authenticity and the non-tangible. An authenticity therefore also disposed to the deformation of reality, which makes “contextual complications transpire […] and where the boundaries between function and representation (which, as in any other text, is extrinsic by way of symbolism, allegory, etc.) are very intricate” (Canella, 1990, p. 5). The objective is therefore the drafting of a map that holds together these contextual complications, interweaving real and intangible data; in our case, this will be done with the method of typological montage applied to the city of Mosul, Iraq. Although in fact the theme of the graphic restitution of certain non-immediately perceptible characteristics of the urban structure and morphology is a theme that can be applied to various realities, we consider it interesting to experiment with and propose this method on a city that is actually and currently in a state of post-war destruction, with confused reconstruction interventions always poised between stylistic restoration and the design of the new without genealogical attention to the place where it belongs. The theme of graphic deformation, or rather the stressing of certain components of the drawing, is considered legitimate by virtue of a long history of cartography that resorts to this expedient for the underlining of certain specific aspects, which will be discussed in the next section.

4 Precedents

Some cartographies will be presented in chronological order; however, this order is in no way intended to assume a linear development of the type of representation adopted. Each cartography will be presented as an autonomous work, relative to the iconology of the time in which it was produced.

4.1 The Peutinger Table

The Tabula Peutingeriana (named after Konrad Peutinger, who first published it in the sixteenth century) is an exceptional cartographic document, a copy made in c. 1200 of a Roman map, datable to around the fourth century AD (Fig. 1). The importance of the document, historiographically speaking, derives from the fact that this is the only surviving cartography, albeit copied, from the Roman period and shows the actual size of the Roman oikumene, extending from England to China. The map resorts to an obvious distortion of the actual geometry of reality as we know it today: a first distortion is immediately attributable to the format (approx. 35 × 680cm), which entails a horizontal elongation and a vertical compression, which, together with the distortion of the angles, makes the map appear unrealistic at first glance. In addition to this format distortion, however, there is a less obvious one, again geometric: the distortion carried out is not linear but reduces some portions of land—where the cartographer had/wanted to show less information—to the advantage instead of the better-known territories, which are more extensive and richer in information. Both distortions (though the former is more exogenous, and the latter more endogenous) are determined by the use of the map, hence its purpose. The map, in fact, does not claim to represent reality, but rather to be a tool for the journey, and therefore to represent the information deemed useful. If the format was designed to be easily rolled up and portable, the deformation of the land is instead attributable to a selection by the cartographer of the main information to be shown, a procedure also familiar to Ptolemy, who “states that if one wants to concentrate the representation of the entire inhabited world in a single map, it is necessary to limit certain parts, which are less important and less inhabited, in favor of others where, as the population thickens, the names and indications would become excessively dense. Moreover, the latter, being better known precisely because of their importance, must be reported by the compiler of a map with an abundance of detail, to the detriment of the space dedicated to the lesser-known ones” (Bosio, 1983, p. 20). This work of selection therefore seems intrinsically linked to the structure of the map, an itineraria picta, and entails the voluntary removal of some detailed information: the seas are not represented with care, being a purely terrestrial map and delegating instead the accuracy to nautical charts (Bosio, 1983, p. 33); the same for “the mountains, the deserts [which] for ancient man constituted areas of minor interest, indispensable for itineraries but not for life” (Bosio, 1983, p. 49). Being instead a continental travel map, mountains assume much more importance for locating the mouths of rivers, which on the other hand are of great importance, “and this can be explained inasmuch as, from the earliest times, watercourses have represented the main routes of penetration, the backbone of traffic, and have given life to a vast and articulated network of communications, which preceded and then accompanied the great overland routes” (Bosio, 1983, p. 55). 55), thus being a consubstantial part of the land routes covered by the map. The most relevant aspect, net of the copyist's errors, is therefore to frame the map not as a summary or imprecise elaboration, but rather to consider the apparent gaps as information that was voluntarily not considered relevant in the drafting, manifesting the editor's awareness of the selection of information and of the objectives the map should have fulfilled, consequently stressing the routes and the ways of communication.

Fig. 1

Segments V and VI of the Peutinger Table (IV century)

4.2 The Map of Europe

The tabula thus introduces an aspect of cartography that is continued into the mediaeval period, where “maps could be shaped and manipulated to meet particular needs as their authors drew from graphic and textual traditions, from experience, and from their own ideas to create individual artifacts suited to given contexts” (Morse, 2007, p. 26). This tendency towards manipulation seems to be justified by the objective that cartography sets itself; the consequences are evident in maps such as the Map of Europe (Fig. 2) by Giraldus Cambrensis, an Irish cartographer, who depicts Ireland, England, and Europe perfectly aligned vertically, to allegorically reduce the distance between Dublin and Rome. If this kind of representation is meant to bring these two territories spiritually closer together (O'Loughlin, 1999), what one can infer is how mediaeval cartographers have “broader concerns, much larger interest in understanding the physical world […that] led them to focus on the system underlying the universe and the laws that governed it. The detail of the earth itself (terra, both the planet and the element earth) were of less interest to them than the grand mechanism of the world (mundus)” (Morse, 2007, p. 30–31).

Fig. 2

The Map of Europe by Giraldus Cambrensis (ca. 1200)

4.3 Opicino De Canistris and Galvano Fiamma Visions of Lombardy

And while the earth often yields relevance to the representation of the mundus and its laws, one finds examples that attempt to combine these two aspects, the real and the ideal, with the conviction to “penetrate into the reality of the world and things in order to know and dominate” them (Morse, 2007, p. 47; Tozzi & David, 1992, p. 348). These overlapping intentions are particularly explicit in the map Pavia and “Lombardy”, dated 1334–1336 by Opicino de Canistris (Fig. 3), in which an allegorical image is combined with a geographical depiction. If on the one hand the course of the Ticino river is represented in an extremely precise manner, in its course and bends—with the exception of the periphery of the drawing in which the course progressively rectifies—the region of Lombardy is abstractly represented as a circular crown around the city of Pavia, within which are inscriptions and descriptions of destinations and localities (Genoa, Emilia, the Alps, etc.), almost as if it were a lenticular distortion, which from a geographical centre fades laterally towards abstraction. (Tozzi & David, 1992, p. 349). Beyond the abstraction of the overlap between the real and the ideal, a further sophistication of representation can be found in Galvano Fiamma. If in the Chronicon maius one can find the ideal reconstruction of a Roman Milan—albeit made ideally circular, it’s probably the first map with an archaeological representation except from Rome itself—in the Cronica extravagans Galvano Fiamma superimposes mediaeval Milan and Roman Milan (Fig. 4). Although the archaeological data is more imprecise (only six gates are represented, described as seven, as opposed to the nine in the Chronicon maius), what is more important is the simultaneous representation, on the same map, of Roman and mediaeval Milan, bringing together in a single drawing an idea of Milan that is as unreal as it is authentic. Galvano Fiamma “dissects the ancient city in its monuments and open spaces, and aims to locate them. Beyond the Milan that flows before his eyes, he reconstructs and relocates the tesserae of the marvellous and blasphemous city of imperial and pagan splendours” for a Visconti Milan “no longer paid only to contemporaneity, but also projected towards the rediscovery of the past fortunes of a city and its protagonists” (Tozzi & David, 1992, p. 352).

Fig. 3

O. de Canistris, Pavia e la Lombardia (1334–1336)

Fig. 4

G. Fiamma, Map of Milan (XIV century)

4.4 G. B. Piranesi and the Campus Martius

If with previous case studies we conclude the iconological parenthesis, with Piranesi we open to an evaluation of the technique of the representation that tends towards invention. To contextualize the Campo Marzio drawings, it seems particularly relevant to point out the differences from the almost contemporary Nolli map—to whose drafting Piranesi himself contributed (Dal Co, 2000, p. 594). If Nolli's map represents a sort of “original truth, […] a literal projection of Rome as it was in the eighteenth century”, and tends to represent the reality of the earth, the “Campus Martius is a web of traces, a weave of fact and fiction” (Eisenman, 2005, p. 40), manifesting instead the presence of that underlying mundus that is not immediately accessible (Fig. 5). Certainly, Piranesi's map offers multiple levels of interpretation: from reflections on the value of architectural language until the ontological doubtfulness of the very possibility of making architecture (Tafuri, 1980); to the general decadence of the discipline from the post-Republican era to the late Baroque (Dal Co, 2000). Nevertheless, it is possible to detect—beyond the polemical aetiology of the map—the graphic possibility of an “urbanism understood as the fabric of memory, rather than as nostalgia for static icons” (Eisenman, 2005), a possible elaboration of the mixture of real forms and types of invention. It is the map itself that, by mixing real and invented data, opens up new considerations, and this is also written by Piranesi himself, in the dedication of the paper to Robert Adam: “Since most of the ruins have completely perished, and others are only small remnants, or almost buried underground, or enclosed in fragments of houses, which it is not easy to discern what they were once used for, I believe that those who wish to further trace the position and form, must certainly be soothsayers, not invented at 'whim', but compared, even in their singularity, by the study of the ancient and the accurate survey of the remains. […] I can truthfully protest to you that there is no such small part of the Campo, which I have not repeatedly and carefully examined […] which remains of buildings, having collected them all, having drawn them with great accuracy” (in Dal Co, 2000).

Fig. 5

G. B. Piranesi, the Campus Martius (1744)

It therefore seems opportune to separate intentions from the result and assess how Piranesi's map lends itself as an operational recovery of architectural history, of the type, with the awareness, however, that “the past offers no teaching and transmits no message if 'curiosity' does not interrogate it and 'invention' does not interpret it” (Dal Co, 2000, p. 599). Far from being archaeologically reliable, according to Purini the value of Campo Marzio is precisely that of demonstrating how Rome—and we could extend this to other urban organisms as well—has within itself infinite possibilities for evoking the ancient (Purini, 2008, p. 31), and how the technique of representation helps, more than unveiling, to interpret and reinvent the traces of history.

4.5 Colin Rowe: Collage City and the Plan Game

With an almost Piranesian inspiration, Rowe and Koetter's work also questions the role of composition through images and references to architecture and buildings from other eras: the more purely “operational” chapters of Collage City (Collage City and the Politics of Bricolage; Collage City and the Reconquest of Time) attempt to find a way out of the stalemate in which, according to the authors, the architecture of modernism is drowning. Regarded as a totalising utopia, in Rowe's dialectic it is exemplified by the Gardens of Versailles (“total control illuminated in full”) and compared instead to the compositional curiosity of Hadrian's Villa at Tivoli, “so disorganised and random, that it proposes the opposite of any totality” (Rowe & Koetter, 1981, p. 148–149).

The text finds in univocity, in the false hope that the future can be naively and deterministically foreseen in the present the great limit of modern architecture, a sterility of meaning that is also recalled in other texts.¹ The objection is that history, outside of a positivistic and naturalistic vision, is not linear, and therefore any architecture that goes beyond its temporal horizon to project itself into the future will always be limited. The alternative that the authors propose, one does not know how polemically, is the “temporary suspension of any tendentially one-sided vision” (Rowe & Koetter, 1981, p. 164), through the structuring of a city composed by collision of fragments, pieces of architecture of various and undetermined extraction. The extreme example is imperial Rome, the same as Piranesi, but the authors identify the same feature in other cities, such as London and Los Angeles. Although the book does not bring any practical experiments of this proposal, the image that accompanies the text is David Griffin and Hans Kollhof's City of Composite, in which it is easy to read the “most tragic awareness of the multiformity of experience, disconcerting and enigmatic” (Rowe & Koetter, 1981, p. 214), once again reaffirming the power of induction over the determinism of deductive procedures. The reference is to Picasso's Bull's Head, with the explanation given by the painter himself: a bull's head formed from a handlebar and a bicycle saddle, which one day, once the artistic meanings have been removed, may once again be used as parts of a bicycle, and not considered art. The form, designated, is thus open to different readings and new meanings, potentially constantly avoiding desuetude and sterility. The most authorial manifestation of this compositional procedure can be traced back to before Collage City, to the so-called Plan Game (Fig. 6) between the Texas Rangers.² John Hejduk describes it as follows: “We would take a large white sheet of drawing paper and take turns drawing plans of real or imaginary buildings. For example, Colin would start with the plan of villa Madama and then Bernhard would connect it with Wright's villa Cage, etc. This went on for the whole night and at first light the sheet was full of plans of buildings from different eras to which many hybrids were mixed. At the end, Colin would study the result with devilish excitement” (in Secchi, 2010, pag. 148). What appears is a random collage technique, the sense of which, however, lies in the selection of fragments and in the inescapable solidity of the memory on which it can be based and from which it can be drawn. A procedure that presupposes “the unexpected coupling of ideas, the discovery of some hidden relationship between apparently remote images; an intelligent expression presupposes an accumulation of knowledge; a memory full of notions that the imagination can conjure up to compose new sets” (Samuel Johnson in Rowe & Koetter, 1981, p. 230).

Fig. 6

The plan game (developed during 50’s period of Rowe in Texas University)

4.6 Saverio Muratori and the Knowledge of Venice

In a certain sense, it seems licit to include in these precedents also the typological cartographic restitutions presented by Saverio Muratori in Studi per una operante storia urbana di Venezia (1960) once the terms of dissimilarity have been clarified. In fact, it is evident how Muratori's entire work represents an organic investigation of the applied case of Venice: from the study of the primitive archipelago settlements to the historical and evolutionary relations between waterways and land routes, to the detailed study of the building fabric. What instead seems to approach the theme of this paper is the path to this kind of knowledge: Muratori writes several times that the objective is the investigation of reality as form (Fig. 7), and the representation of form in drawing is such an important result that it is not compromised by any inaccuracies and approximations of the surveys conducted by the students on his course. Indeed, research is presented as an intentional process—conditioned and oriented, in a positive sense—that has within it a hypothesis to be verified: “reality would not be identifiable and isolable without the choice of constants, that is, without an intentional sense of research. […] And this knowledge is concrete to the extent of its capacity to translate the subjective intentional vision into the positivity of reality. Hence knowledge is a vision-reality relationship, the positivity of which is in the degree to which a hypothesis of value adopted as a constant reference and total vision succeeds in producing, that is, in interpreting and containing the real” (Muratori, 1960, p. 9). History, considered a founding aspect of knowledge/operativity, is therefore the condensation of an experiential and cognitive baggage that is poured into the drawing, which gives an account of the type-morphological features of the building fabric, and allows us to rediscover the internal order of coherence and structural succession of the city's development; but it is also the legitimate—precisely by virtue of history—reconstruction of reality in a conceptual way (Muratori, 1960, p. 10).

Fig. 7

A table extracted from Muratori (1960)

5 Methods and Results

The reconstruction of reality in a conceptual way, and consequently by the inductive route, proves to be an extremely relevant method of knowledge for those case studies, such as the city of Mosul, where the conceptuality cannot be replaced by observation of reality. In this way, even though the war's destruction has compromised much of the building and monumental fabric, one can inductively rediscover and reconstruct that trace—which Eisenman defines as a non-absent absence—that profoundly characterizes the urban identity and remains latent in the urban fabric. In this, it is preliminarily necessary to know some typical characteristics of Islamic architecture and the city, which can create a scientifically reliable vocabulary for the conceptual reconstruction of the city's image. To this end, a typo-morphological investigation of the city through the redrawing of a conceptual map, several residential case studies were selected—four from Mosul and nineteen from Baghdad, extrapolated from historical and documented surveys.³ On these selected cases a redrawing and typo-morphologiucal analisys process was then carried out (Fig. 8), highlighting certain common characteristics—such as the structuring around a central courtyard; the presence of iwans and talars (Fig. 9); the general introversion of the building structure.

Fig. 8

Redrawing of several residential units

Fig. 9

Highlighting of two invariant features: the iwan and talar disposition (left) and introversion (right)

From the redrawing—as well as from the more general study of the structure of the Islamic city—an attempt was made to abstract some possible type-morphological invariants of the urban fabric, including.

• The introverted character of the building, both residential and monumental: the settlement attitude of the Islamic world is in fact articulated on the presence of one—or more—internal courtyards, source of light and air for all the rooms of the building. This courtyard often takes on definite and precise geometric connotations, opposite to a multilinear and irregular building perimeter.

• These courtyards and/or patios act as collective open spaces, compensating a densely populated urban structure.

• The courtyard and/or patio structure provides for a tendentially horizontal density of the building fabric, which is generally no more than three storeys above ground.

• The scalarity and typological hierarchy between residential fabric and primary elements: the structure of the Islamic city can be assumed as a matryoshka representation. If the entire city, often walled—at least originally—can be recognized as an enclosure with the mosque-souq system at its centre, so too the internal cul-de-sac ramifications lead to “neighbourhoods” with the mosque and a minor souq as their centrality. This situation imposes a hierarchy of courts: from the largest city mosque in relation to the city; to the local mosque concerning the neighbourhood; to the residential court in terms of residential units.

• The Mosul area—but extendable to Baghdad as well—is influenced by certain Iranian architectural components, such as the iwan, an emptying of the inner perimeter wall of the courtyard, which tends to have a pointed or round arch vault.

To better exemplify the procedure followed, reference will be made to a portion of fabric above the mosque of Nabi Djirdjis (Fig. 10). Starting from an aerial photograph—a photograph from 2012 was chosen, to be able to assess an urban fabric not compromised by war destruction and in its vital state—a redrawing of the area was carried out, highlighting its fundamental components: the perimeter of the streets that outline the blocks, the identification of the internal courtyards, and the definition of the perimeters of the residential units. The result is a map of solids and voids, a sort of rooftop plan from which it is clear how the urban structure is essentially characterized by the punctual presence of primary monumental elements, around which spontaneous residential fabrics have developed, almost disorderly in their appearance as a solid mass excavated only by the open spaces—streets and internal courtyards.

Fig. 10

The survey of solid and voids

To continue the investigation from a typological point of view, two reference parameters were chosen as constants: the street layout—both in its aspect of defining the blocks and as the distributive character of the cul-de-sac—as permanence; and the presence of courtyards as an invariant (Fig. 11). Starting from these two geometric data, we then proceeded to the progressive selection of the most suitable residential case studies—from those previously redrawn—and their assembly within the perimeter of the residential unit obtained from the aerial photo, giving priority to the size of the courtyard and the correction of any incongruity with the street profile. For timing reasons, and not considering this bias detrimental to the demonstrativeness of the operation, priority was given to the typological assembly on the street fronts, leaving the centre of some blocks undefined instead (Fig. 12).

Fig. 11

Example of procedure to adapt residential case studies on the context starting from courtyards and streets

Fig. 12

The montage process

By combining the assemblages of the various areas, the result is a typological map (nominally 1:1000) of the innermost part of Mosul's historic centre, an area of extreme importance that stretches from the Al Nouri Mosque to the citadel area on the Tigris River and the bridge. The map arbitrarily mixes elements that are still present (the Dominican complex of Al Sa'aa, the upper part of the Souq), and elements that have been destroyed (Al Nouri Mosque, Nabi Djirdjis Mosque, the lower part of the Souq) with imaginary—though fundamentally plausible—elements introduced through montage. The aim was also to emphasize the typological difference of Nineveh Street from the historical fabric, in which there are mostly courtyardless mansions with commercial units at the street level, built at the time of the carving of Nineveh Street from 1944 onwards. The typological combination, while stressing some specific aspects of the urban structure, enhances certain type-morphological features and reveals a possible sense of interpretation of the city's form, condensing an interpretation of Mosul's structuring characteristics into a design that is not susceptible to time (Figs. 13, 14).

Fig. 13

The virtual reconstruction of Mosul Old City

Fig. 14

Detail of the previous map

6 Theoretical Precedents and Evaluations

At this point, it seems opportune to carry out an evaluation of the whole operation, which presupposes two fundamental questions: what is being represented, and whether this representation can play a cognitive role. It is interesting to quote some passages from Rossi (1970) on Venice: “Venice […] is particularly interesting because it proposes a set of organized questions that go beyond its urban reality; they constitute an idea of Venice”. An idea of Venice that abstracts the mere real status of the urban facts, and intersects it with the myth, the perception, and the memory of the city, with the aim of constructing an analogous ideal image of Venice itself. The construction of an analogous alter-ego of the city increases the visibility of certain specific and typical characters, which might otherwise disappear in the historical remixing of the city's physical material. “This observation interests me to introduce some questions on architecture in the very sense of this research. Take the question of historic centres: I believe that the most serious way to work on cities, or to understand them, which is not very different, is to mediate between the real city and the analogous city. […] In fact, I ask myself, is there a historical centre of Venice that is authentic or actually represents any historical section of the city? No one could argue this for Venice or for any ancient city in Europe. […] Those who have taken the question of historic centres seriously have ended up reducing the real city to a moment in the city's history and compressing the urban dynamic within these forms; roughly what is called environmental or stylistic restoration. […] The real alternative is to proceed with the construction of the analogous city” (Rossi, 1970, pp. 424–425). If, for Rossi, the term construction identifies a phase probably subsequent to the one we are considering—the actual phase of planning—the reflection on the authenticity or corruptibility of the real artifact of the city is extremely relevant. The construction of an idea of the city, a city with recognizable features, passes through the abstraction of some of its ideas (or, as Quatremere de Quincey would say, images) that are able to grasp certain aspects of urban memory, and that enhance them.

In his essay Houses of Memory (2014, pp. 221–236), Eisenman quotes Freud's famous passage on Rome, in which the psychoanalyst writes that “if we wanted to represent the historical sequence in spatial terms, we could only do so by means of a juxtaposition of space: the same space cannot have two different contents”. The historical sequence—or rather, the psychic history Freud refers to—can be, according to Eisenman, divided into two categories—as mentioned at the beginning—those of history and those of memory. If the domain of history prevails as long as the form and function of the architecture remain tied to their original intent, when the function decays and the form resists one enters the domain of memory, of the succession of events. Memory thus contaminates the built environment with those traces, invisible but perceptible presences-absences, that make up the idea and image of the city. And if Freud fails in “superimposing” these two instances—the physical and real one of history with the psychological and oneiric one of memory—according to Eisenman's interpretation of Rossi, the meeting place between these two issues is precisely the creation of an analogous space; an authentic space in which takes consistency the relation between the abstract idea and the mnemonic features of the locus, thus escaping the risk of utopia as a non-place. “But unlike the city, the urban skeleton, the analogue is separated from a specific place and time, to become instead an abstract locus that exists in what is a purely typological or architectural time-place. […] The place of the analogy is thus abstracted from the real city. It is a non-place, yet it is a different place from the modernist utopia precisely because it is rooted in both history and memory. […] Here the analogous city turns out to subvert the real city. Whereas the skeleton was read as the form of the city's specific time and place, the design process of the analogue displaces the city's specifics of time and place in favor of another reality, a psychological reality based on memory.” (p. 231).

If the analogue, then, can imply a condition, albeit paradoxical, of reality, how many analogues can there be? How many representations? The multiplicity of the result is, however, a guaranteed condition that separates the construction of an analogue from the construction of an ideal. In the preface of the book Palladio Virtuel (Eisenman, 2015), Eisenman attempts to define the limbic condition of a virtual state. It is neither the “literal” construction nor its ideal. Referring specifically to the title, Eisenman writes how Palladio has redrawn his buildings not as they were actually constructed, but as Palladio himself wanted them to be known; and that, therefore, the reality of Palladio's work should be contextualized in the space between the drawings and the buildings, thus as a virtual Palladio. What interests us, however, is an earlier passage in which Eisenman writes: “The architectural ideal refers to an organization of form. The virtual refers to architectural relationships implied by a condition of presence that exists beyond the literal and the ideal” (p. 10). The literal as actual, the ideal as an external reference, almost Platonic in its fixity, the virtual remains a mode of reading, explanation, that moves between these two extremes, and lends itself to the world of variable interpretation. As also in the analyses of Palladio's villas that are later proposed in Palladio Virtuel, there is not just one virtual, there are several virtuals, each highlighting a different characteristic within the same ideal. When we speak of virtual as reality, as existence, it is inevitable that we also refer back to the trace, to the memory mentioned earlier. In this space of abstraction, therefore, the operation of montage seems to find legitimacy, which does not risk the inaccuracy of other attempts of summary survey—inevitable in the effort to reproduce an urban state of affairs—and does not claim the ideality of a univocal reconstruction, instead aiming to restore a specific latent interpretative space of a condition that is both real and immaterial.

7 Conclusions

The result thus obtained is a map of a virtual Mosul that is of decisive importance in understanding and reading the city, as well as an essential part of its design definition. The approach, and the consecutive selection, of certain features of the city and their material relocation on a map, not only seeks to understand reality as a form—that is, to relate individual facts and relationships back to the overall form of the city—but also “to understand a series of constituent processes of urban reality […] and to understand why these processes have produced certain forms” (Aymonino, 1977, p. 12). Indeed, the stress placed on certain characteristics chosen as constants—permanencies, invariants…—allows a synoptic vision of the urban phenomenon, to grasp the hierarchies between spaces and the type-morphological relations that have led to the urban form and, ultimately, to the graphic restitution of the sense of the city's places. And in this sense, it stands in continuity with certain cartographic modes presented previously: distortion as emphasis, the representation of the mundus instead of the earth—with its substratum of meaning and the laws implied in it—and above all the manifestation of an abstract condition of the city deliberately taken as a hypothesis to achieve a cognitive reconstruction of the city by way of concept.

Beyond the built material destroyed by the war and beyond the metamorphic capacity of the residential fabric—always changing, as well as always stable in its constitutive laws—what remains is the representation of a conceptual Mosul in its own and identity-based characters. The map lends itself to being an instrument of investigation, suspended between the reconnaissance and interpretative condition of a survey, but also the limiting case of a possible design conformation. And if urban reconstruction cannot, evidently, find a match in the mimetic process of slavishly remaking the old, but needs the evolutionary and transformative awareness of the city's structure, this tool is placed between the instances of identity and modification, as a synthetic act between knowledge and the inductive capacity of the project to emphasize certain traits of authenticity, albeit within an evocative and conceptual framework.

8 Future Developments and Methodological Implications

The effort expended here on the drafting and theoretical contextualization of what we have called the representation of a conceptual Mosul implies, however, the opening of a further disciplinary counterpoint. If on the one hand it has been appropriate, up to now, to strongly delimit the montage operation in a strictly reconnaissance and cognitive sphere—to leave no doubt about its instrumental use—it is also inevitable to recognize in it the potentialities of an acerbic transformative reasoning. The limit, hitherto imposed as an uncrossable boundary, is actually a variable spectrum that superimposes the reading of reality on a plausible design interpretation. The possibility of grasping the value of this gap is already implicit, for example, in the table of the Analogous City (Fig. 15), proposed by Rossi at the 1976 Biennale di Venezia, in which real, imaginary, and design elements are mixed to additively create a conceptual horizon broader than just perception and logical comprehension, attempting to touch the chords of the relationship between reality and imagination (Rossi, 1976, p. 5). And when the evocation of a generic irrational hypothetical runs the risk of not leading to any verifiable conclusion, here is where memory can instead interpret that non-logical design impulse that shelters architecture from the risk of being merely a container flattened on its own function. In the miscellany of references mounted on the board, the most explicit one regarding the transposition from memory to project may be the juxtaposition of a typological plan in which the imprint of a Roman orthogonal settlement modality is clearly legible, with the residential complex proposed for San Rocco in Monza (in competition with Giorgio Grassi, 1966). The assemblage here reveals both the part of memory, the centuriation, and the part of imagination, the residential complex, which in its form recalls the territorial structuring of the Roman matrix in Lombardy, as well as the settlement identity of one of the most important public projects of Renaissance Milan, Filarete's Ospedale Maggiore. And if on the left side of the table we have the premise and on the side the result, the juxtaposition represents the process of conceptual transformation of a trace that becomes a project.

Fig. 15

La città analoga by Aldo Rossi

The interposition of a graphic tool—the reference—as a mediation between the subject, the designer, and the object, the project, thus takes on the appearance of a theoretical-conceptual contrivance that condenses in a typical element both its compatibility with history and memory (precisely by virtue of its own ancestry and origin) and, due to its intrinsic function as a reference, the possibilities of being modified, adapted, and redesigned. One cannot fail to notice how the recognition of this role of the type corresponds in design, if seen explicitly, to the technique of montage. Thus, “a new apparatus-object makes its appearance: an object—as non-subject—that for the first time both analyses and invents” (Eisenman, 2014, p. 230). To invent, meaning its Latin root of invenio, to find the right link between form and context, between the role of the significant of form and that of its content. Guido Canella also moved in this direction when he responded to a Casabella call for design proposals in 1979 and submitted a project for the Centro Direzionale area of Milan (Fig. 16)—what has now become the Porta Nuova Project. Given the short time available to formulate the proposal, Canella relied on previous, unrealized projects by architects linked to the city of Milan—Terragni, Pagano, but also Leonardo da Vinci. The meaning is, however, more sophisticated: the decision to use the montage of some unrealized works has as its objective to reconstruct the functional-formal identity of a sector of Milan; to find those authorial projects that, on the one hand, self-expose themselves as a sort of museum of Lombard Architecture, but also manifest the invariant and structural relationship between architecture and the city, to “connect and contextualize operationally and critically the living and suspended cornerstones in the building body of the city [… from which] result the historically invariant features of a local culture that has always been originally aimed at formalizing the settlement between experimentation-functional innovation and tradition-formal regeneration” (Canella, 1979, p. 102). The most interesting contribution, insofar as it is extremely didactic, is precisely the explication of the tension between the first design phase—preliminary to the modification—and the analysis of the invariant characteristics of the city's structure, all through a graphic representation entrusted to the references.

Fig. 16

The montage panel by Guido Canella

And if Rossi faces the problem from a point of view both personal and analogical, and Canella instead resorts to a more solid demonstrative process rooted in context, attempts of this processualism can also be found—albeit on a more morphological scale—in contemporary projects, such as Peter Eisenman's Yenikapi Archaeological Museum in Istanbul (Fig. 17). Here, the project has three levels of invention/analysis with respect to the memory of the context: the first is undoubtedly the reconstruction of the trace of the Theodosian Harbour, from the Byzantine era, which corresponds with the physical structure of the museum; the second concerns the morphological and settlement trace of the Yenikapi district: the rotation of the urban layout—manifested by the two directions of the railway line—is reaffirmed with the configuration of two perimeters, rotated and overlapped, which impose themselves as a support plane and background to the archaeological figure of the museum; the third concerns the characterization of the inner distribution system, since the geometry of a reticular mesh found in the mosque of St. Sophia is superimposed on the project. Beyond compositional virtuosity, it is evident how the summation of elements always moves in the balance between unveiling and adding, in the invention of a new, yet authentic, narrative heritage. If the results of these three design procedures, as already mentioned, certainly have substantial differences, one cannot fail to emphasize the intentio implicit in these attempts. Memory is seen as an operational tool of composing, which has not only an additive task—the addition of the new—but also a subtractive one—the manifestation of the constant, of the invariant of the underlying urban structure. All this is condensed in the identification of an intermediate body, the architectural precedent, which mediates between the instances of design and memory, an intermediate body that legitimizes with objectivity a historical and transformative process already imprinted in its genetics, but which also preludes to its future possibilities of modification.

Fig. 17

Yenikapi archaeological Museum’s drawings by Peter Eisenman

Admitting, therefore, the legitimacy of recourse to the precedent—as well as for the reconstruction of reality by conceptual means set out earlier—also for design practice, it is considered interesting to try to decline these reasonings also on the application case, Mosul, and to attempt an evaluation of operational procedures compatible with this context. The application area is the district of Maiden (Fig. 18), which was completely destroyed following the bombardment suffered by the regular armies stationed on the eastern front of the Tigris River. The study of the context begins with the same assumptions as the typological assemblage: recognizing in the morphological street conformation a permanence and in the introverted courtyard structure the main invariant of the built-up area. Also noteworthy in this neighbourhood is the presence (now absence) of the Shaik al-Shatt Mosque, historically among the most important in Mosul (Herzfeld & Sarre, 1920).

Fig. 18

The Maiden district

The procedure, which is still to be understood as a pure research direction and not as a final result, began with an analysis of the surface area considered, highlighting the relationship between solids and voids (Fig. 19), from which one can see, in addition to a clear preponderance of the built-up area, how the open space is mainly articulated by private courtyards. In the second instance, an attempt was made to understand the edges of the area (Fig. 20), dissecting its box-like structure: just as the neighbourhood is visible as a single built mass in a first phase, two primary thoroughfares cut through the area. Despite this, the distributional structure mostly develops in the inner and secondary streets, thus identifying five built-up agglomerations. Following this analytical phase, a critical observation of the urban structure of the agglomeration was conducted (Fig. 21). Attention was paid to the indications of orientation and measurement: the alignment of the buildings tends to be oriented perpendicularly to the river (as do the internal courtyards), and it was shown that it is possible to identify a measurement (albeit approximate) that seems to be a reasonable estimate of the depth of the built-up area: the measurement of 8.5 m is in fact appropriate to represent the limit of the buildings and the median line of the courtyards.

Fig. 19

The area

Fig. 20

The permeability of edges

Fig. 21

The structure: modules and measurements

An attempt has therefore been made to conceptualize this settlement modality, for a synoptic representation of these structural components of the city. Although still lacking in development, this mode of approach is considered extremely promising, as a comparison with previously redesigned case studies shows that the measure identified is indeed compatible with the size of residential units (Fig. 22). The study proposed here, which is still in the conception phase, is nevertheless similar to the methodologies adopted in the previous analyses: if assembly, drawing, and the selection of invariant features have proved to be essential tools for reconstructing the virtual idea of the city, to give it a clear and exhaustive representation, it seems equally promising to use such methodology also for the development of preliminary tools for the project that attempt not to reconstruct the formal appearance of the context, but to investigate its underlying logics and authentic and implicit metrics.

Fig. 22

The measurements of residential units