In some cases in Sindarin, an initial [s] before a consonant could become syllabic and develop into [es-], such as ✶skalā > S. esgal “cast shadow” (PE17/184), ✶spini- > S. esbin “thin thread” (PE17/17), or √STEN > S. esten “short” (PE17/184). All of the attested examples of this development involve an initial [s] followed by a voiceless stop.

Conceptual Development: In the Gnomish and Early Noldorin of the 1910s and 1920s, it seems syllabic [ṣ] was part of the normal phonetic inventory of Primitive Elvish, although it was rare. In Gnomish it seems that [ṣ] became [us], at least before labialized consonants: ᴱ√ṢQṢ > G. usc or usg “fog” vs. ᴱQ. usqe (QL/98, GL/75). In the Early Noldorin of the 1920s [ṣ] became [is]: ᴱ✶sṣt- or ᴱ✶pṣt- > ᴱN. hist “spits” (PE13/147, 163). This may have been the normal development of [ṣ] in the Early Period when it did not precede a labialized consonant (we see the same in Early Qenya). In the Ilkorin language of the 1930s, we also see that initial [s] became [es] before voiceless stops, which may be the inspiration for this Sindarin sound change.

In some cases in Ilkorin, an initial [s] before a consonant could become syllabic and develop into [es-], such as ᴹ✶spálass(e) > Ilk. espalass (Ety/PHAL) and ᴹ✶skal(a) > [eskal >] Ilk. esgal (Ety/SKAL¹). All of the attested examples of this development involve an initial [s] followed by a voiceless stop. In Tolkien’s notes associated with the Comparative Tables, he said (PE19/27):

Beleriandic partly shared this early tendency [of simplifying initial clusters] but in general shows its early connexion with Telerin in hesitation in this matter (producing both f and esp from initial sp)

In Tolkien’s notes on the development of the Beleriandic mode of Tengwar from the 1940s, he said (PE22/31):

in Doriathrin ... est etc. was the normal form taken by old PQ initial st

Both the quotes indicate that the normal development was [s] into [es] from an initial s + voiceless stop, though perhaps not universally. One possible counter-example is Ilk. istil “moon” which might have developed from *ṣtil < ᴹ√SIL. If so, the quality of the vowel that developed from the syllabic [ṣ] may have depended on the following vowel.