2.8.1. 1. The so called mobile s refers to the phenomenon of alternating word pairs, with and without s before initial consonants, in stems with identical meaning. Some examples are
(s)ten- thunder, strong noise - lat. tonare, ohg. donar, osl. stenjo.
(s)pek- look carefully, spy - lat. spectus, ohg. spehon, alb. pashë
(s)ker- cut - lat. caro,
(s)qalos- big fish - lat. squalus, ohg. whale
*(s)teros- bull - gr. lat. tauros (from a cognate word), ger. steer.
NOTE. Some think it was a prefix in PIE (which would have had a causative value), while others maintain that it is probably caused by assimilations of similar stems (some of them beginning with an s-, and some of them without it). It is possible, however, that the original stem actually had an initial s, and because of phonetic changes (due probably to some word compounds, where the last -s of the first word assimilates with the first s- of the second word) it was lost by analogy in other situations. This would explain why in some languages both stems are recorded, and why there seems to be no pattern in the losing of initial s-, as each word would have had its own history in each language.
2. Before a voiced or aspirated voiced consonant, s was articulated as voiced, by way of assimilation. So in nisdo, nest, or misdho, salary. When s forms a group with sonants there is usually assimilation, but this trend is sometimes reversed by adding a consonant (as in lat. cerebrum, from /kerəsrom/).
3. The s between vowels is the most unstable, with different outputs depending on the dialects. Example of this is snusos, daughter-in-law (lat. nurus, ohg. snur), but the most common ones are roots endings -s with a declension beginning with a vowel, as in osl. nebesa (gr. nefeos), cloud, lat. generis (gr. geneos), lineage, or got. riqizis (gr. erebeos), darkness.
NOTE. In Germanic, as in Latin, the intervocalic -s- becomes voiced, and then it is pronounced as the trilled consonant, what is known with the name of rhotacism.
2.8.2. A sequence of two dentals is usually eliminated in all Europaio modern languages, but the process of this suppression differs among languages. It is supposed that this trend had already begun at the time of the IE II, and so Europaio probably had already some phonetic changes in these consonant groups - possibly still common to all dialects. As we have said before, in phonetics we prefer to sacrifice the search for purity for the sake of unity, and so in this case we look at the earliest situation possible. So, in forms such as ed-tos (lat. esus), weid-tis (lat visus) or sed-tos (lat. sessus), we won't usually represent any phonetic change, unless we take the word as a derivative loan word.
2.8.3. The manner of articulation of an occlusive or sibilant usually depends on the next phoneme, wether it is voiced or voiceless. So, for example, the voiced ag- carry gives voiceless gr. aktos, or lat. actus; the same happens with voiced aspirate, as in legh-, gr. lektron, lat. lectus, ohg. lehter; voiceless p- becomes -b for example in zero grade pod-, foot, which appears as -bd-, as in gr. hepibda.