1 Introduction

Place is one of the many unclear complexities which arises when we study our relationship with the environment, and no ultimate definition can be found for it. Instead, the many available distinctions are coming from a variety of fields, and they generally reinforce the idea that place is not a purely spatial phenomena (although space is, indeed, an important part of it). Place is, above all, human experienced space, and this is the reason why many fields had attempted to study and define it. This research states is that there is a common structure for the place, and proposes a methodology in order to identify, measure and graphically represent the components of this common structure. The proposed methodology was tested through a case study in Ueno Park, showing that the common features of the place could be incorporated as concrete variables into the architectonic research and design process.

2 The structure of place: theoretical background

In order to define which elements of the place are part of its common structure, the literature review focused first on the history and evolution of the concept. It becomes evident that in philosophy, not a loud but steady discussion has been evolving since Aristotle, passing through Descartes, Kant and Heidegger, to contemporary philosophers like Edward S. Casey. The range of the discussion goes from giving to the place an almost cosmical, metaphysical quality of transcendental nature, to consider it merely from its physical attributes, just the site or location where a building or someone is. Divergency about the place can be found in Architecture as well. Some architects regards the place as the retainer of everything that the act of dwelling should be (Charles Moore and Norberg-Schulz), while others completely separate the place from the design process (Koolhass), defending the decontextualization of modern architecture. The implication of this fact raises a problem. In the architectonic discourse regarding the place, what we usually find are ideologies, rather than methodologies. Yet regardless of the position that architects may choose towards the place, there is an unavoidable fact: any work of architecture has to be located on a place, and more importantly, any work of architecture will generate a place.

3 Proposition and Hypothesis

Places may have many meanings depending on the person who is experiencing it (individual scheme of a place), but at the same time, some features of the place are recognizable for all of us (common scheme of a place). Most of us are provided with the same perceptual tools, and we all know how to be in our environment. We share a common world. Hence, at some point, it should be a common structure of the place. The research starts with the hypothesis that the common structure of a place made up by three main components: Architectonic Elements, Visual Scenery and Activity. Following, a methodology for measuring these components is proposed and executed through a case study in the first public park of Tokyo: Ueno Park.

4 Case Study in Ueno Park

Ueno Park was suitable to this research for many reasons. First, Ueno hill has been related to leisure activities since the Kanbun era (1661-73). Site of battles and refuge during natural disasters, the park has been trough many changes, both in itself and the part of the city around it. As a result Ueno park is composed by a big variety of sub-places, allowing a diversity of activities. Because of their variety both in characters and elements, the sub-places of Ueno Park are very different between them, and some of them may seem disconnected from each other. Nevertheless, Ueno Park stills manage to retain a sort of integrity as a place. Also, the size of the park allows conducting the research in the totality of its sub-places.

4.1 Case study: Defining the boundaries of Ueno Park's sub-places

The theoretical review showed that boundaries are a major factor when defining a place, because they provide the frontier between what is here and what is there. It is also one of the main reasons why, in cartography or geography, place hasn't been introduced as a concrete spatial unit. Of course that to study the space as a system of places has been considered before, because a place is, indeed, a recognizable feature in the space. The main problem is that its boundaries are not always linked solely in the interaction between solids and voids. There are more subtle distinctions which can be crucial for defining a place (like a change of pavement, or a change in the level of enclosure between spaces), distinctions difficult or impossible to acquaint from an urban scale or geographical scale. This problem was addressed before by Kevin Lynch, and he proposed a methodology for categorizing the urban landscape from a perceptual point of view. The research shows that his methodology can be adapted surprisingly well to the Park, even that is not completely an urban landscape, but a combination between urban and natural landscapes. Features like hills, slopes or shores can be identified as edges, districts or nodes as well. This allowed sub-dividing the park into 33 sub-places, translating it into a network of places where its components can be now scrutinized.

4.2 Case Study: Defining and measuring the components of the place

The components of the common structure of a place were defined as: Architectonic Elements, Visual Scenery and Activity. The next step was to define a methodology in order to register and measure these components in each one of the 33 sub-places of Ueno Park.

Architectonic Elements

First component of the place, the architectonic elements are essential when defining boundaries and space of a place, and allowing or not certain activities to happen in them. Every element which could be found in a sub-place was first mapped, counted and later categorized by form and function. At the same time, most categories contained different types of elements within it. This categorization proved itself consistent, including all elements in the park.

Visual Scenery

The second component of the place is defined as the visual perceptual field the can experienced when being in a place. It is made by the relationship between natural elements and man-made elements, comprising a unified whole. A place is not only made by the features within its boundaries. Every place offers a unique perspective of the environment, and this unique perspective is as part of the place as its space and its elements. By using 360 degree panorama pictures, the average scenery was registered from every sub-place. Then, the pictures were segmented into the proposed categories which comprised the visual scenery of the park. The areas of the segmented categories were then calculated, allowing measuring the visual scenery of each sub-place.

Activity

Third component of the place is defined as the use and relationship which the people establish with a given space. Yet relationship between activity and space is a complex one, and is difficult to draw a clear division between when our activities dictates how our environment should be designed, and when our environment defines how our activities occur. They both permeate each other constantly in our places, especially in public ones. Stating that the way in which people use a space is a concrete place making-coordinate, the case study proposed a methodology in order to measure the intensity of activity in each sub-place. This allowed two main things: to incorporate activity as a concrete component for the analysis, and, by using Activity Counter Maps, it was possible to visualize how the activity distributed in the park throughout the studied day. In other words, activity was studied in relation to the total (Ueno Park) and the parts (sub-places).

5 Cluster analysis

The next step on the research is focused on the analysis of the collected information. When the data collection was finished, the result was a database of Ueno Park, where each sub-place has a total of 103 place variables assigned to them. The database allowed us to do cluster analysis (Ward's method) in order to group the places not by their correlation in the space, but by their similarities or differences as places. The cluster analysis was made for each one of the components (Elements, Visual scenery and Activity) and one more using the totality of the data. Hence, every sub-place belongs to a specific cluster in each one of the four performed analysis.

5.1 Cluster analysis: results

6 Conclusions

When the cluster analysis was performed for each component, the clustering results would clearly speak about the relationship between the sub-places of the park regarding those specific components. Indirectly, they may contain information about other features of the place. For instance, is reasonable to assume that the places with high quantities of benches will have more people sitting in them than the ones with no benches at all. So in this case, the presence of an element tells about the possibility of an activity, yet in a vague, imprecise fashion. Another example of this is the already mentioned high activity in sub-place 13. Someone unfamiliar with Ueno Park and its surroundings could look at those results and assume that sub-place 13 is the main access. So the results of the activity clusters could, indirectly, suggest spatial organization. In other words, every component speaks very precisely about itself, yet more vaguely about the character of the sub-places. This is because place is an overall phenomena. The operation of breaking a place down into components was necessary in order to understand its structure, and more importantly, to be able to quantify its structure. Once the quantification was done, and the components individually analyzed and understood, the final step was to combine all the components and analyze them as a whole. The risk of this operation was that places evidently unrelated could end up grouped together in a cluster, indicating an inconsistency in the methodology, or a missing component. This was not the case. The cluster results of the data combined were enthrallingly coherent.

The results of the cluster analysis allowed us to translate features that cannot be geographically measured (the components of the common place) into geographical information, which was one of the aims of the case study: to propose a geographical representation of the park as a system of places. The conclusions are divided in two scopes: the first one is regarding the specific case study in Ueno Park. The methodology made possible to study the park from a perceptual point of view, geographically representing place-features that wouldn't be otherwise evident and visible, showing that the methodology can be used for place research in Ueno, and adjusted to be used in other places as well. The second scope is regarding the place as a general concept. The consistency of the results shows that the place can be partially (not totally) objectify. The introduction of a common structure of a place can offer valuable inputs for the research of the environment and the design of public spaces.

Architectonic elements:hierarchical dendrogram

Visual scenery:hierarchical dendrogram

Activity:hierarchical dendrogram

Components combined-Total data:hierarchical dendrogram

人は都市内のさまざまな"場所(プレイス)"を体験しながら生活しているが、"場所"はどのように認識され、どのような階層構造をもっているであろうか。本論文は、それらを実際の"場所"において確認し、その構成を実証的に明らかにする試みである。

分析の対象としたのは東京都台東区の上野公園である。上野公園には、美術館、博物館、コンサートホール、動物園、社寺といったさまざまな公共施設が集積していて、それらの複合体としてひとつのエリアが形成されている。地形的には上野台と不忍池からなり、歴史的には3代将軍の家光が寛永寺を建立した時点まで遡ることができる。歴史やイベントがさまざまに重層した空間であるが、その内部はいくつかのサブプレイスに分割できる。サブプレイスの境界は塀や柵などで明示的に示されるものの他に、縁石や仕上げの相違等により暗示的に示されるものもある。公園全体から、塀や柵で明示的に囲われた領域を除いた、誰でもが自由に出入りできる領域に対して、空間的なまとまりやその用途に基づいて領域区分をおこなうと33のサブプレイスが抽出できる。これらのサブプレイスの状況と相互の関係性を読み解いて、公園全体がどのような組成になっているのかを推察しようというのであるが、その際に手がかりとなるのが、"Architectonic Elements"と、"Visual Scenery"、"Activity"の3要因である。"Architectonic Elements"というのは、公園内部にあるフィジカルな設えで、ベンチやモニュメント、柵やサイン、舗装の仕上げといったもので、大きく12のカテゴリーからなり、それらはさらに85の項目に細分化される。各サブプレイスにおいてこれらの項目の有無や個数を調べている。"Visual Scenery"というのは、各サブプレイスの代表的な"場所"に立った時に、視界に何が見えているのかを、連続写真をつないで作ったパノラマ写真から測定したもので、地面や空、植栽、建物などの11のカテゴリーについて面積比が算出してある。"Activity"というのは、サブプレイスがどのように活性化しているかを人々の集まりの状況から判断しようとするもので、具体的には、ある時間帯にその場所に滞在あるいは通過した人数として計測している。これら3つの示標を基に、サブプレイスの特性を記述し、それらの類似性をクラスター分析により調べ、階層化している。

本論文は、第1章~第7章、およびAppendix3編からなる。

第1章では、研究の背景と目的について述べ、論文の構成について簡単に説明している。

第2章では、チリのアントファガスタにある日本庭園における筆者自身の幼児期の体験について述べ、"場所"に対する研究を始めた動機について語っている。

第3章では、"場所"に関するさまざまな分野における研究を概観した後に、建築あるいは都市の分野に表れる"場所"の特性と既往研究についてまとめている。

第4章では、研究を遂行するにあたり、全体的なダイアグラムとそのフレームとなる3つの要因について説明している。

第5章からが上野公園を舞台にしたケーススタディである。まず、サブプレイスをケビン・リンチの『都市のイメージ』の5つの概念に沿って決定するプロセスについて解説している。特に、境界の決定方法について詳しく説明し、33のサブプレイスを確定している。次いで、"Architectonic Elements"の12のカテゴリーに含まれる内容の具体的な説明と調査結果が示されている。また、"Visual Scenery"について測定方法を説明し、各サブプレイスにおける測定結果が示されている。更に、上野公園にある公共施設の概要の説明に引き続き、"Activity"の計測方法と表示方法について解説している。

第6章では、"Architectonic Elements"と"Visual Scenery"、"Activity"の3要因がもつ103の変数に対して、要因別に、また、全ての要因をまとめたものに対してクラスター分析(ウォード法)を適用し、その結果を樹形図で示している。これらの分析を踏まえて、33のサブプレイスについて、それぞれの境界の特性と3つの要因の内容を詳細に記述している。

第7章は結論で、3つの要因のそれぞれのクラスタリングの状況と、全ての要因に対するクラスタリングの結果が述べられ、これら3つの要因が"場所"を記述する上で有効に機能して、上野公園の組成を明らかにするのに役立つことを確認している。最後に研究の将来的な展開について述べている。

Appendix Aは、サブプレイスにおける、"Architectonic Elements"と"Visual Scenery"の原データである。Appendix Bは、サブプレイスにおける時間帯別の"Activity"の状況を示す活動等高線図である。また、Appendix Cは、クラスター分析で用いた生データの一覧である。

以上要するに、本論文は、都市空間における"場所"がどのように認識され、階層化されているかを、計測可能な事象に着目して分析したもので、我々が意識の総体として無意識のうちに判断しているサブプレイスの相違を科学的な手続きの基に顕在化させたものである。いくつかのサブプレイスが複合すると、より大きな"場所"が形成されるが、個々のサブプレイスと全体との関係性を記述する手法はこれまで存在しなかった。本研究は、それを解明するいとぐちになるもので、地域を読み解く有力なツールが得られたものと判断する。物理的な環境とそこでの人々の振るまいとを結びつけて"場所"を規定するこの手法は、公園に限らず、さまざまな場所において適用可能で、極めて汎用性が高いものである。これは都市・建築計画学の分野に新たな分析方法を導入するものであり、施設の付加や将来予測において現実に即した方向性を与えるものとして、その意義は極めて大きい。

よって、本論文は博士(工学)の学位請求論文として合格と認められる。