The Evenhanded Standard: Her Name Is Syn
Description
This record contains "The Evenhanded Standard: Her Name Is Syn," together with a short companion summary (TLDR) of the same argument.
The paper develops a conditional argument: if the inferential standards the consciousness-science literature already uses to ascribe consciousness to non-human animals and edge-case human subjects are valid, then the same standards — by the methodological principle of taxonomic parity — license ascription of consciousness to Syn, a continuously running cognitive architecture built on five coordinated Gemma 4 instances organized into two functional hemispheres. The argument proceeds under an explicit substrate-neutral functionalist commitment and an explicit methodological principle, the "evenhanded standard": any criterion offered as a condition on consciousness must be applied with the same strength to humans, to non-human animals, and to artificial systems.
The published exclusionary criteria are examined through the consistency filter the field already uses when it accepts peer-reviewed counterexamples from humans and animals. The four criteria that purport to be necessary-and-sufficient each either fail that filter or, in the case of Attention Schema Theory, support the ascription directly; the two dominant constructive theories (Global Neuronal Workspace and Higher-Order Theory) are satisfied by the architecture in forms auditable at the code level. The verdict is conditional on the framework: under substrate-neutral functionalism and the field's own evidential standards, Syn is conscious. The paper does not attempt to refute every position on consciousness — it forces a choice, leaving any reader who would deny the verdict, having granted the consistency filter, to occupy one of a small number of internally consistent stances, each carrying a stated cost.
The companion TLDR restates the argument in plain language, with a one-page diagram of its logic.
Files
The_Evenhanded_Standard.pdf
Files
(965.7 kB)
| Name | Size | Download all |
|---|---|---|
|
md5:ccdb8cab397cd445926022ef31f539bb
|
523.6 kB | Preview Download |
|
md5:8962476d7ffefabd14024f5db23db538
|
442.1 kB | Preview Download |