觀察、不充分決定與理論評價

本文旨在表明，實用論對「理論」與「觀察」關係的看法，不會傷害「競爭理論可客觀地比較優劣」的主張。實用論的「觀察理論」及「經驗資料不充分決定理論」之論題分別是：觀察已經預設某些東西，無法成為驗證或否證理論的中立裁判；理論若遭遇觀察異例，可透過修改其中任一部份使之與觀察相符而免遭反駁。這兩種觀點可能導致相對主義或懷疑主義。本文將批判此立場並論證，觀察者本身的各種預設，固然會影響觀察陳述的客觀性，但由此不必然導出競爭理論無法客觀比較優劣的結果。待檢驗的理論偶爾可以透過修正其中部分，使之與觀察異例相符，但不是所有理論遭遇觀察異例，皆可透過修正而與競爭理論等價並獲得保留，我們能以既有證據和各理論比較的方式，獲得一個相對較佳的理論。

關鍵字

科學哲學；觀察；負載理論；不充分決定；迪昂；貝耶斯主義

並列摘要

This paper aims to show that the pragmatistic theory of observation and thesis of the underdetermination of theory by data do not damage the centre notion of rationalism, namely there are cognitive reasons to evaluate rival theories and decide which theory is the better one. The pragmatistic theory of observation holds that all observation is theory-laden and so does not provide a neutral court of appeal for assessing the cognitive value of theories. The underdetermination thesis implies that any theory can be maintained in the face of any evidence or of all possible data, provided we make sufficiently radical adjustments elsewhere in our beliefs. I critique them both and argue that the case for 'theory-ladenness' preventing observation from testing a theory is somewhat overstated, and the difficulty of underdetermination of theory by data may be damaging to early rationalism but not to later rationalism which is more moderate.