A unitary, immoveable, ‘Brugmannian’ Proto-Indo-European was developed for decades, where all differences between branches were attributed to dialectal exceptions in the vocalism of the parent language. That concept was changed for another one, represented by the widespread acceptance of a ‘laryngeal’ Proto-Indo-European—thanks especially to the decipherment of Hittite.
However, the simplistic view—already present more than seventy years ago—of a unitary, abstract, atemporal parent language, from which all other branches would have split at the same time, has changed little. The field has changed one simple concept by another, slightly more correct. But the main error remains: immobility.
Phonetics seems to be often the subject of change in the field: first the satem-centum distinction, then to shared isoglosses, then from vocalism to laryngeals, including the gradual acceptance of the archaic nature of Anatolian.
With this paper, we propose that what is often described as infinite independent events of laryngeal loss, intertwined with multiple independent exceptions, be exchanged for general rules of stepped laryngeal loss, coupled with a reasonable number of exceptions for each dialectal period.