Graziano et al. (in press) sketch some suggestive ways in which Graziano’s attention-schema theory (AST) of consciousness (Graziano, 2013, 2019; Webb & Graziano, 2015) connects with three other theories: the global-workspace theory (GWT) of Baars (1997) and Dehaene and Naccache (2001), higher-order-thought (HOT) theory (Rosenthal, 2005), and illusionism (Frankish, 2016). According to Graziano and colleagues, these “theories should not be viewed as rivals, but as partial perspectives on a deeper mechanism” (1). The deeper mechanism they have in mind is that of AST, which they urge constitutes “a kind of standard model that covers the basics if not the details” of consciousness (1). I’ll raise difficulties for this irenic construal of these four theories, and argue that they are in important ways theoretical competitors. Graziano and colleagues see the term “consciousness” as applying to two distinct phenomena. One is “how select information reaches a state in which it is bundled centrally and can impact output systems.” They call this i-consciousness, in contrast with what they call m-consciousness, “a more mysterious, extra, experiential essence that people claim accompanies the informational content” that occurs in i-consciousness (2). The “m” of “m-consciousness” reflects their view that this notion is inherently mysterious. With that distinction in place, Graziano and colleagues urge that the central issue for a theoretical treatment of consciousness is “whether m-consciousness, the more ethereal notion of consciousness that people intuitively believe they have, is accurate or instead is an imperfect model of i-consciousness” (2). I’ll raise concerns about both these notions. To begin with, i-consciousness is not something we would naturally classify as any kind of consciousness at all. I-consciousness is exclusively an informationprocessing notion, and all the relevant processing can occur without being in any way conscious. Indeed, Graziano and colleagues themselves refer to i-consciousness as merely “the content of consciousness” (2), and the content of consciousness is typically something that can occur either consciously or not. As Graziano and colleagues note, GWT is in effect “an account of i-consciousness” (2). So the foregoing is also a difficulty for that theory. GWT holds that a state is conscious if its content reaches a global workspace, making it available to downstream processing across various modules. But being globally available is not itself a way of being conscious. There is doubtless considerable overlap between the states we classify as conscious and those that are globally available for downstream processing. But even if that overlap were total, there would be no reason to see being globally available as a way for a state to be conscious. And the overlap is far from total. Visual states remain conscious well into the periphery, however scant in detail, and conscious peripheral states are rarely available for global processing. Similarly for conscious background auditory sensations. And unconscious thoughts and desires sometimes have global effects. These considerations also apply to Block’s (1995) notion of access consciousness. I-consciousness and access consciousness are not types of consciousness at all. Graziano and colleagues hint at this when they note that GWT “leaves unexplained how people end up believing they have a subjective experience” (13). Their oblique formulation, about “people... believing they have a subjective experience,” is because they see subjective experience as m-consciousness, which they see in turn as mysterious and ethereal. But the

中文翻译：

---

意识的竞争模型

格拉齐亚诺等人。（正在出版中）概述了 Graziano 的意识注意力模式理论（AST）（Graziano，2013 年，2019 年；Webb 和 Graziano，2015 年）与其他三个理论的联系：巴尔斯的全球工作空间理论（GWT） 1997）以及 Dehaene 和 Naccache（2001）、高阶思想（HOT）理论（Rosenthal，2005）和幻觉主义（Frankish，2016）。根据 Graziano 及其同事的说法，这些“理论不应被视为竞争对手，而应被视为对更深层次机制的部分观点”（1）。他们想到的更深层次的机制是 AST 机制，他们敦促它构成“一种标准模型，它涵盖了意识的基础知识（如果不包括细节）”（1）。我将对这四种理论的这种冷酷的解释提出困难，并认为它们在重要方面是理论竞争者。Graziano 及其同事认为“意识”一词适用于两种不同的现象。一个是“选择信息如何达到集中捆绑并影响输出系统的状态。” 他们称这种 i-意识，与他们所谓的 m-意识形成对比，“人们声称伴随信息内容的更神秘、额外、体验的本质”发生在 i-意识 (2) 中。“m-意识”的“m”反映了他们认为这个概念本质上是神秘的。有了这种区别，格拉齐亚诺及其同事敦促对意识进行理论处理的核心问题是“m-意识，人们直觉地相信他们拥有的更空灵的意识概念，是准确的还是不完美的 i -意识”（2）。我将提出对这两个概念的担忧。首先，我意识根本不是我们自然会归类为任何一种意识的东西。我意识只是一个信息处理概念，所有相关的处理都可以在没有任何意识的情况下发生。事实上，格拉齐亚诺和他的同事们自己将 i-意识仅仅称为“意识的内容”（2），而意识的内容通常是可以有意识或无意识发生的东西。正如 Graziano 及其同事所指出的，GWT 实际上是“对 i-意识的一种解释”（2）。所以上述对于那个理论来说也是一个难点。GWT 认为，如果状态的内容到达全局工作区，则状态是有意识的，使其可用于跨各种模块的下游处理。但在全球范围内可用本身并不是一种有意识的方式。毫无疑问，我们归类为有意识的状态与全局可用于下游处理的状态之间存在相当大的重叠。但即使这种重叠是完全的，也没有理由将全球可用视为一个国家有意识的一种方式。而且重叠远非全部。视觉状态在外围保持有意识，但细节很少，并且有意识的外围状态很少可用于全局处理。同样对于有意识的背景听觉。无意识的想法和欲望有时会产生全球影响。这些考虑也适用于 Block (1995) 的访问意识概念。我意识和通达意识根本不是意识的类型。Graziano 及其同事在指出 GWT“无法解释人们最终如何相信他们拥有主观体验时”暗示了这一点 (13)。他们关于“人们……相信他们有主观体验”的斜文表述是因为他们将主观体验视为 m-意识，而他们又将其视为神秘而空灵的。但该