Chapter 6. Critical methodological pluralism --- Intensive and extensive research design
Combining methods
Critical methodological pluralism 批判的方法論多元主義：
メソッズは多元的であるべき（様々な性格のものがあるべき）だが、方法間の適不適は頑として存在する。その適不適はメタ理論とメソッドの関係で決まる。［→前段はごく当たり前の話だが、どの辺が新しいかといえば、メタ理論の部分だろうか？］
We would like to designate this attitude critical methodological pluralism, in which the foundation for what is suitable or not is to be found in the relationship between metatheory and method.
プラグマティスト的方法論選択に対する批判 共感にみせかけてのdisり。存在論と認識論は（プラグマティストが思ってるより）ずっと大事だよというこの本の通奏低音。
Howe holds that epistemology must not be placed above practical issues, nor the conceptual over the empirical. This is also a fundamental starting point in the pragmatic perspective. The practical and the empirical take precedence over the ontological and the epistemological, a view that the pragmatists themselves call ‘the dictatorship of the research question’ (Tashakkori and Teddlie 1998).
Although we sympathize with this pragmatic attitude to the issue, we would like to emphasize the importance of paying attention to the ontological-methodological link. We too want to see more methods in use --- when necessary. However, there is a great risk that some conclusions will be drawn that cannot be drawn from the application of a particular method unless you have made the ontological base clear. (p.152)
pp.153ff: いま流行りの混合研究法は批判的方法論多元主義とは似て非なるものだよという話がずっと続く。要するに、混合研究法において量と質の協調関係・統合と称されるものは、多くの場合、存在論的・認識論的考察を欠いた「協調・統合」であるという批判。
Quantitative method
量的手法は、論理実証主義的なメタ理論に依拠していることが多いので、まず論理実証主義とは何かを考える。批判的実在論的な言葉遣いをすれば、「私たちの知識の基礎は経験的ドメインにある」と考えるのが実証主義。
あと、方法論的個人主義。
Qualitative method
質的手法のメタセオリーは様々。解釈学、現象学、etc.
現象学の微妙な位置。客観的経験を重視するという点で実証主義的だが、全体への志向、および、内省の（経験的事実以上の）重視という点では、反実証主義的。
In one sense phenomenology is positivistic and in another sense anti-positivistic. Among the positivistic features is the belief in an objective method (free from all prejudice) and the belief that sense experiences are the source of knowledge (proximity to empiricism). A pure science based on ‘pure’ facts and not on idle speculations is desired. What primarily distinguishes the phenomenologists from the positivists is their different perspectives: phenomenologists are seeking the whole and not, like many positivists, the parts. Another difference is the view held by the former that even though sense-data are the primary source of knowledge, it is by reasoning that we attain a deeper insight into reality ? through introspection. In this process consciousness must be cleansed from the researcher's subjective thoughts. One can achieve this by letting doubt occupy the seat of honour. Phenomenology can therefore be characterized as close to empiricism and rationalistic; it takes its starting point in ‘simple sense-data’ and by means of pure reasoning it reaches true knowledge. (pp.160-161; emphasis added)
Intensive and extensive research design
量的 vs. 質的という伝統的区別に代えて、インテンシブなデザイン vs. エクステンシブなデザインという区別を提案。
インテンシブなデザイン
- 因果的な集団を研究し、
- 諸現象の実質的な関係に注目する。
- メカニズムの因果的説明を目指す
エクステンシブなデザイン
- 分類的な集団を研究し、
- 諸現象の形式的な類似関係に注目する。
- （説明的でない）記述的な一般化を目指す（ただし準規則性の検出には役立つ）
Relationships and types of groups: what constitutes a social phenomenon?
インテンシブ／エクステンシブで、焦点をあてる集団の性質が決定的に違うという話。
インテンシブなデザインでは、因果メカニズムを説明するために、因果的集団（実質的な関係を持つ集団）を対象にする。
一方、エクステンシブなデザインでは、その（サンプリング上の）特性上、分類的集団（単に類似カテゴリに属するだけであり、実質的関係があることが保証されていない集団）を対象にするため、実質的関係を見つけることは難しい。
Working procedures and sampling: why is contextualization important?
エクステンシブなデザインにおけるサンプリング
→ 母集団への推測のため
インテンシブなデザインにおけるサンプリング
→ 事例研究的サンプリング。(1) 極端な事例を見つけてくる戦略。(2) 大きな変動を含む事例。(3) クリティカル事例（理論上逸脱している事例）。 (4) 標準事例
Some methodological restrictions
伝統的な「統計分析であれ数学的モデリングであれ数字では複雑な社会的現実は理解できない」という批判。そして、それに対する反論「数字による抽象化＝複雑性の縮減こそが、社会的現実に肉薄するための道具だ」
↑こうした不毛な（？）「質 vs. 量」論争へのコメント、３点
その１「単純化はいずれの立場であれ不可欠な作業」
First, probably nobody would deny the fact that social reality is so complex that some form of simplification is necessary to enable scientific studies. It is therefore misleading to criticize simplification as such, arguing that social reality cannot be simplified but that the complexity must be maintained if one is to understand the phenomenon. People inspired by postmodernism often take such a view. They are extremely sceptical about such theories of which we gave examples above, i.e. rational choice theory. In particular they are critical of the grand philosophies like Marxism. However, we argued in Chapters 2 and 3 that simplifications are both necessary and fruitful, and in Chapter 4 we discussed in great detail modes of procedure for this, by means of abstraction. (p.172)
↑ポストモダニズム的研究プロジェクトを、経験的データ（or 具体的事例）の分析と接続するときの矛盾も指摘されている。
その２「複雑性とは開放システムに由来する問題である」
開放システムでは分析対象・メカニズムが絶え間なく変動するからこそ、複雑さが生まれてくる
その３「モデルの認識論的基盤に関連付けて議論しないと不毛」（本書がここまで何度も言ってきたクリシェ）
量的研究は、研究プロジェクトのあるレイヤーにおいては不可欠。
In spite of the criticism we relate here, we maintain that quantitative analytical models can be fruitful and are undoubtedly valuable ? provided that their field of application is confined to what is suitable. It is of great value, for instance, to be able to measure and register changes in welfare development, so-called social statistics ? measurements often calling for sophisticated mathematical methods. We have emphasized before that quantitative descriptions are often indispensable in the research process. The limitations are revealed mainly when it comes to explanatory ambitions. It is vital to be very observant of what kind of data one uses in the analysis. Since it is generally formal relations of similarities and taxonomic groups that are involved in the extensive approach, one can hardly draw any conclusions of a causal nature from such analyses. One must be careful not to overrate the method, and be aware of its limitations. (pp.174-175; emphasis added)
要するに、量的研究は、測定に使うのは有用、説明のために使うのは望み薄（なぜなら「エクステンシブ」では形式的関係しか取り扱えないから）。
Types of result
インテンシブ vs. エクステンシブ の違い、まとめ
Several times already we have commented on the different types of result which are the outcome of the two approaches, and it should have become clear that the crucial difference is that it is via the intensive approach that one is able to expose generative mechanisms. In an intensive design one can (at best) explain a particular occurrence or a particular object, as well as a larger social phenomenon. The extensive approach alone cannot answer questions of causation; on the other hand it can point out important empirical manifestations of mechanisms. Thus it can provide vital descriptive information, which may be useful as support in a discussion about causal powers, and it can contribute to generating questions of causality. In the fundamental issue of the relationship between the specific and the general, both intensive and extensive designs are indispensable features of the research process. (p.175)
Conclusion
略。
