User:Mellohi!/sandbox/Kortlandt effect

The Kortlandt effect is a sound law for Proto-Indo-European where the phoneme d, traditionally reconstructed as a voiced coronal stop, is debuccalized into the laryngeal h₁ in certain environments, especially before other consonants. It is named after Dutch linguist Frederik Kortlandt, who initially conceived of the sound law in 1983.

History
In 1983, Frederik Kortlandt, while collecting support for his version of the glottalic theory, formulated the sound law to account for irregularities in the evolution of Indo-European terms for multiples of the powers of 10. Most saliently, the term for 100 was previously reconstructed as *ḱm̥tóm, but the Ancient Greek descendant ἑκατόν has an initial ἑ-. Before Kortlandt, it was assumed that this initial epsilon came from prefixed *sem- "one" due to its initial rough breathing, which regularly arises from word-initial *s-. This in turn requires the irregular disappearance of the prefix-final nasal. Kortlandt instead proposed that the term really started with *h₁ in Proto-Indo-European, noting that ε- is noted a common reflex of word-initial *h₁ before a consonant in Greek. The rough breathing would thus be secondary and analogical. The Kortlandt effect allows for the term for 100 to be more firmly derived from *déḱm̥ "10" via *dḱm̥tóm, with a change from *d to *h₁.

Alexander Lubotsky found more examples of the Kortlandt effect in 1992 and 2013.

Romain Garnier coined the term "Kortlandt effect" (effet Kortlandt) in 2014 to refer to changes from *d to *h₁. Garnier notes further examples of the sound law applying when the *d precedes a consonant.

Examples
The following Proto-Indo-European formations are described as being affected by the Kortlandt effect.

*deḱ- "to take, perceive"
Vedic Sanskrit dāśvāṃs- "devout, pious" has long been derived from "to take, perceive". Unlike most Sanskrit formations of its kind, akin to the perfect participle, the root vowel seems to be lengthened instead of the root being involved in reduplication. The Kortlandt effect allows a derivation from a reduplicated participle formation *dedḱwos, from which the second *d would be transformed into *h₁ and thus cause the apparent lengthening in Sanskrit.

*dwi- "two, apart"
The Sanskrit adverb and prefix "apart" may be derived from  "two" via the initial *d turning into a laryngeal and then disappearing word-initially before a consonant.

*h₂éd "at, to"
Romain Garnier explains the hitherto unknown etymology of the Sanskrit preposition  with the Kortlandt effect. He identifies Latin ' "at, to" and Germanic, both from Proto-Indo-European ', as cognates. To account for the long vowel and loss of the final consonant in Sanskrit, Garnier proposes that *h₂éd alternated with an extra form *h₂éh₁. The latter form with an extra laryngeal appeared before consonants, while the form with a *d appeared otherwise. Indo-Iranian would then generalize *h₂éh₁ while the other Indo-European languages generalized *h₂éd.