Norwegian monosyllables
Consonantal circumvocalic dissimilation
Hans Vogt, in "The structure of Norwegian monosyllables", Norsk tidsskrift for sprogvidenskap, 12 (1942): 5-29, at p. 22 says about »Government between final and initial consonants« [in Norwegian monosyllables:] »In general it may be said that such government is inexistant. Consonant "harmony" in one sense or the other is unknown. There are, however, some cases which point to such rules«.
The sequence rær occurs only when the final r is morpheme, e.g. in træ:r plural of tre:, 'tree', ræ:r plural of rå, 'pole'. »This leads to the discovery and formulation of a more general rule: The same sonant does not occur in the initial cluster and after the vowel. Words of the type pri:r, skle:l, fli:l, vra:r do not exist - with the exception mentioned above.« (i.e. when -r is a morpheme.)
You can have C¹VC¹ (same consonant before and after the vowel), for example lal:, lul:, lel: (< li:kevel:), but not C°C¹VC¹; only exception is the word pleil. »The same rule seems to be valid for initial n clusters, apparently incompatible with post-vocalic n or n(g).«
Vogt seems, during WW2, not to have read Twaddell's article from 1939-40, "Combinations of consonants in stressed syllables in German". It sounds contradictory when he states that the l, r and n are not repeated (in C² and C³) in Norwegian monosyllables of the form C¹C²VC³, but that it applies only to these three consonants. Apparently he only looks after rules that are fully implemented ("laws"), and thus he overlooks trends that are just as real, but having a different distribution than the "rules". Claiming the avoidance of repeating three of the Norwegian consonants (and only them) in circumvocalic position gives more cause for wonder than to understanding, while the statement that C¹VC¹ tend to be rarer than you would expect statistically, tells of a general function of the language; and we must expect that this is precisely the case. The rule of the three consonants (l, r, n) becomes a special case of the tendency.
At about the same right, one could argue that there is no rule for the formation of the plural of English nouns, as any text is full of words like thief - thieves vs. chief - chiefs, mouse - mice vs. house - houses, die - dice vs. lie - lies, and plural forms as men, women, children, brethren, oxen, feet, teeth, geese, mice, lice are rather frequent words. Another group is sheep, deer, bird, also fish, giraffe; dozen, million, and often also foot. Radius - radii, phenomenon - phenomena, crisis - crises too are quite common types, in certain kinds of texts. Nevertheless, it is generally assumed that English plural nouns end in -s (maybe the rule might better be described as a tendency).