When I lived in Germany ('75-'76) older faculty at Georg-August Uni called what we call a-umlaut the "German A" so I doubt Arnold's claim that Kant would be [kaent]; I suggest it's a short variant of what the LANE phonetics used the "upside down a" for.