The concept BUTTERFASS, especially the so-called STOSSBUTTERFASS, is referred to by numerous geosynonyms.

Some names have such striking phonetic similarity in the stem that their connection is hardly questionable:

(1) roa. pigna, with the tonal vowel variants [ɪ, e, ɛ, a], etc.;
(2) sla. pinja, an obvious romanism as its area of distribution is contiguous with that of (1);
(3) roa. pinacc, which is a suffixed form of (1);
(4) roa. panaglia (with variants of the initial unstressed stem vowel, corresponding to those mentioned in (1)). This type is dominated by variants with the unstressed stem vowel [a]

(5) The type pignatta 'pot', also known in Standardita., can be assigned to (1) along with its frequently masculine dialectal variant (cf. AIS 973), too. Although it is more commonly documented with the meaning 'pot of terracotta' (cf. AIS 955) within the VA area, outside of it, namely in Emilia-Romagna, it explicitly denotes a pot in which smaller quantities of butter are churned by beating (with a wooden spoon, etc.) (cf. the legend in AIS, Karte 1206, Type C).

Morphologically and semantically, it is thus logical to think of a designation for a vessel pigna as a base type. The type of designation – VESSEL FOR CHURNING – also speaks for such a base type as a term for the general conceptBUTTERFASS:

latte di pigna BUTTERMILCH, literally 'milk from the churn' (in Trentino).

Of factual interest is that the STOSSBUTTERFASS, which appears archaic, is not the oldest technique, as its specified Romansh designations panaglia lunga, literally 'long butter churn', and panaglia dret sü, literally 'upright butter churn' (Lower Engadine), show (cf. AIS 1206).

However, the origin proposed for ita. pignatta to ita. pigna 'pine cone' (< lat. *pīnea[m]) – "prob. [...] per la somiglianza di forma delle più antiche pignatte con una pigna" – is not convincing semantically. It is true that the conical shape of some terracotta and bronze pots may be reminiscent of pine cones (cf. DELI) but a historical reference to the subject that is decisive for the history of the word can be taken from the AIS map mentioned above 955 LA PENTOLA (PIGNATTA) DI TERRACOTTA. In fact, it also contains a list of names for BRONZE POTS (AIS 955_2), some of which have been transferred to it secondarily, especially in the Alpine area, since they go back to a completely different material used to make cooking pots, namely the so-called soapstone, ita. steatite, laveggio, deu. also Lavetz(stein) (cf. the commentary on AIS map 963, LA MARMITTA as well as AIS 970 IL VASO PER LO STRUTTO). This versatile material, which was comparatively easy to use because of its rather soft surface was mined mainly in the Ticino and Lombard mountains and was also used to make other objects, such as stoves, which in Romansh are also called pegna, (Engadine) pigna (HWdR, 571; LRC, 798; on pigna, pegna 'oven of soapstone' cf. the commentary on AIS 937. These ovens, incidentally, are 'nearly cubic' (AIS 937, commentary)) and do not bear the slightest resemblance to a pine cone.

Consequently, this is a clear case of metonymic polysemy (rather than homonymy). Pigna 'stove' and pigna 'vessel for churning butter' are named after the material from which both were made: soapstone. However, it is not strictly necessary to assume a pre-Roman etymon, as Alexi Decurtins suggests in LRC, 798) for Romansh pegna | pigna 'oven'. Formally, the etymology could well be that proposed by G. B. Pellegrini *pinguia (Latin pĭnguis 'fat') – but not elliptically from pinguia(m) (ollam) in the sense of a 'vessel (= Lat. olla) for fat' ("Recipiente particolare per conservare il grasso, fosse esso strutto, sugna, o burro cotto, oppure un arnese elementare per fare il burro" ([1976, p. 171 cit. DELI 928]), but in the sense of a mineral or rock similar to fat in terms of its appearance and consistency (cf. similarly motivated deu. Speckstein). Latin *pinguia (petra) 'soapstone' is therefore suggested as the base type for (1)–(5).

The many forms with the stem vowels [ɐ, a] show a strong and obvious onomasiological influence by the etymologically sepate panna 'cream'. Not related to this type, however, are

(6) Lombard pench, raw. paintg 'butter', which should be traced back directly to pĭnguis 'fat' (HdR).

(7) Roh. penn 'buttermilk'

could be a regression based on pigna 'butter churn' since the buttermilk is drained from it.

The following diagram shows the word family (green arrows) as well as the recorded meanings (red arrows).

With regard to the metonymic motivation of the polysemy, the transfer of the terms from the natural raw material to artefacts of increasing complexity made from it (simple vessel > mechanical device) and finally to the functions associated with it can thus be observed.