Imposing a shevu'a is one of the many instruments available to Beit Din to process disputes. Often, the very threat of a shevu'a and the specter of violating one of the Aseret Ha-dibrot by issuing a false oath, are sufficient to deter litigants and elicit confessions. Generally, oaths are Rabbinic in origin only three have Biblical roots: modeh be-miktzat - an oath for someone who partially concedes a claim; shevu'at eid echad - for someone whose denial is contravened by one witness; and shevu'at shomrim - the oath taken by a shomer who hasn't successfully secured the item. All other oaths are purely Rabbinic in nature.
Having established an eid echad as a force which can generate an obligation to swear, we must now ask whether one witness also exempts a person who has been obligated to swear. By empowering an eid to obligate shevu'a, is the Torah essentially assigning to one witness full reliability pertaining the issue of shevu'a? Regarding monetary matters the testimony of two witnesses is required for us to believe a claim, but for oaths - whether to obligate them or to exempt them - one eid suffices. A gemara in Bava Metzia (2b) is highly suggestive that, indeed just as a single eid can obligate a shevu'a, so may he exempt one. The first mishna in Bava Metzia assesses a case of conflicting claims of ownership, when two people struggle over an article of clothing or claim they purchased the identical item from the owner. The disputed item is divided among the contestants, each of whom must take an oath (Rabbinic in nature) to support their claim. The gemara questions, "Why not consult with the seller and inquire to whom he sold his item?" Tosafot interpret the gemara's question as follows: Shouldn't the testimony of the seller, if not sufficient to award the item (since he is only one witness), be adequate to exempt one party from his oath? Why should each party swear, when the seller's testimony might function as an eid echad to replace one party's shevu'a? According to Tosafot, this query (which is not fundamentally rejected) models the halakha that one eid can exempt shevu'a.
Two lines of reasoning are suggested to explain this rule, which is known as 'eid echad mesayeia.' The simple approach is to suggest that the eid replaces the shevu'a. Being that an obligation to swear constitutes a responsibility to clarify the situation, one witness' testimony might function in lieu of that oath. If the personal oath would have been sufficient clarification, then certainly the testimony offered by one eid should also suffice. In this respect, the eid echad substitutes for the shevu'a. By contrast, we might claim that the eid echad helps the litigant avoid the shevu'a. The presence of eid echad does not serve as a substitute, but rather cancels the obligation to swear in the first place. (Admittedly, this might depend upon the particular shevu'a which eid echad seeks to exempt. Exploring the mechanism which generates a shevu'a would be paramount towards our understanding of how an eid echad might inhibit that very mechanism.) In summary, and to employ classic Talmudic terminology: is the eid echad "bi-mkom" ha-shevu'a (REPLACEMENT), or is an eid echad a "poter" (EXEMPTION) for the shevu'a?
An interesting nafka mina is articulated by the sefer "Sha'ar Ha-mishpat" – a commentary written on Choshen Mishpat. The halakha of gilgul shevu'a allows secondary claims – which in their own right do not obligate shevu'a – to be incorporated in a primary shevu'a. For example, if a dispute occurred surrounding land, no shevu'a is obligated. If a subsequent debate between the same two parties about cash did generate a shevu'a, the original dispute over land can be incorporated into the current shevu'a. "Gilgul" literally means "rolling," and it allows non-shevu'a disputes to be integrated within a case subject to shevu'a.
What would happen if a structure of gilgul would be arranged, and subsequently an eid echad exempted the original shevu'a? Would the tovei'a (prosecuting litigant) be allowed to plead the added gilgul shevu'a even though the base shevu'a has been dissolved through eid echad? Presumably, this question would stem from the fundamental understanding of eid echad's efficacy. If the eid echad substitutes for a shevu'a, the supplementary shevu'a would still be pending and can be pursued. If, however, eid echad revokes the very basis of the original shevu'a, any supplementary shevu'ot, built upon, and expanded from, that basis would naturally be annulled.
A second question might pertain to a type of shevu'a known as nishba ve-notel - a shevu'a preceding collection. All Biblical shevu'ot are of the variety known as "nishba ve-eino meshalem" - defensive shevu'ot taken by the defendant to reinforce his non-payment. Certain situations which favor the defendant still require an oath to fully corroborate his position and sustain his innocence. Many Rabbinic shevu'ot were fashioned along these lines. However, the mishna in the 5th perek of Shevu'ot lists several individuals who may take an oath which facilitates subsequent collection. These oaths are known as nishba ve-notel (literally, 'he swears and collects'). Would an eid echad be effective as a replacement for a shevu'a which enables subsequent collection? A sefer known as "Teshuvot Maimoniyot" (a collection of Ashkenzic responsa addressing various issues in the Rambam), in siman 61, suggests that eid echad would, in fact, facilitate collection without shevu'a. (See as well the Mordekhai to Bava Metzia, siman 219, who cites the Maharam MiRotenberg as asserting this position, and the Rema, Choshen Mishpat 87:6, who cites this opinion as a 'yesh omrim.')
Conceptually, it is more likely that an eid echad would NOT obviate a shevu'a which enables collection. Especially if one sees the eid chad as canceling the basis of the shevu'a, collection should not be enabled. The most coherent manner to explain this shita might be to view eid echad as a replacement for the shevu'a; by summoning one eid, the litigant has produced something equivalent to shevu'a and in place of shevu'a. As such, he might collect without actually swearing, since he provided an equivalent replacement to the required oath.
An interesting comment by the Rosh might shed light upon the efficacy of eid echad mesayeia. The gemara in Bava Metzia (36) addresses a situation in which an appointed shomer transfers the item to a second party for guardianship. According to Rava, the original owner can indict the first shomer by claiming, "I trusted you to watch and swear if you could not return the item, while I do not trust the second party's oath." The first shomer's inability to provide a reliable oath obligates him to pay. Now if one witness' testimony can substitute for required oaths, why should we obligate payment from the original shomer because he cannot provide an oath – his oath should be replaced by the testimony of the second shomer!!! Evidently, at least according to the Ramban, testimonies or eid echad do not replace oaths!!!
The Rosh cites this issue (Bava Metzia 1;3) and defends the eid mesayeia concept by claiming that even though eid echad can exempt an oath, it cannot exempt payment required if a person is unable to swear. Very often, people unable to properly swear must pay the actual demand. In the case of a shomer who transferred an item, his inability to swear (since the item was no longer under his guardianship) and the unacceptability of the second shomer's oath, yield an obligation to pay. One eid's testimony will not excuse this obligation. Apparently, the Rosh viewed the effectiveness of eid echad as replacing an oath. One witness may replace an oath, but certainly cannot excuse a debt which emerged as a result of inability to take an oath. By contrast, had eid echad annulled the very NEED for an oath, the subsequent inability to take a non-required oath (due to eid echad) would not have generated a monetary obligation.