The gemara in Gittin presents the famous takana of Hillel enacted in order to allow debts to remain un-cancelled by shemitta. As the gemara explains, Hillel was concerned that the specter of imminent shemitta would discourage lenders who would fear the dissolution of their debts. He therefore established the safety valve of pruzbul which, as the gemara explains, provides 'relief' (pruz is Aramaic for assistance) for both rich (buli = rich in Aramaic) and poor (buti=indigent). The rich can lend with security and the poor benefit from this sense of security and the consequent generosity.
The gemara does question the capacity of Hillel's takana to override the Biblical shemitta. To this the gemara responds that Hillel's takana was only fashioned for periods during which shemitta is derabanan (certainly our day and, according to many Rishonim, even during the days of the second Beit Ha-mikdash). The gemara does not, however, directly address the mechanism established by Hillel. Possessing the authority to override shemitat kesafim derabanan, what mechanism did Hillel employ to facilitate this process?
An interesting Sifri to Parashat Re'eh suggests one such mechanism. The pasuk forbids demanding debts after shemitta by writing "you shall rest your hand from that which your brother possesses [of yours]." From this the Sifri infers that yours which are not fully possessed by "your brother" (since they are already partially collected) may be collected even after shemitta. From this the Sifri infers that any loan to which a collateral (mashkon) was assigned may be collected after shvi'it and is not subject to shemitta cancellation. By securing a mashkon, the creditor has already 'begun' the preliminary stages of collection (since if the debtor defaults he will just keep the collateral) and hence is unaffected by shemitta. The Sifri continues, "Hillel used this principle as a source for the establishment of pruzbul." The Sifri indicates that Hillel borrowed from the mashkon mechanism in establishing his takana. Ideally, only a mashkon may deem the loan partially collected. Hillel institutionalized this phenomenon by allowing the enactment of pruzbul to serve as an approximate mashkon. By entering his 'demand' in court and initiating the collection process, the creditor has mimicked the situation of mashkon. Tosafot (Gittin 36a) cites this Sifri (albeit in a slightly different version) linking pruzbul to the mechanism of mashkon.
An interesting gemara in Gittin implies this version of pruzbul. The gemara claims that a pruzbul may be written only if the debtor owns land - even a minute amount. The Rash in Shvi'it (10:6) explains that the land condition is fundamental to the pruzbul's efficacy, as it contributes to the condition of initiated collection. Inasmuch as land is not portable and entails the primary source for collection, the creditor already holds a legal lien upon it. This partial kinyan – known as shibud - creates an approximate state of initiated collection similar to mashkon.
Rashi in his comments to an earlier gemara in Gittin (32b s.v. mosrani), provides an alternate definition to pruzbul. The Torah forbids a creditor from prosecuting the collection of his loan ('Lo yigos'). Writing a pruzbul is the equivalent of appointing beit din as the principal collectors of the debt. Once they are authorized to collect, the violation of lo yigos does not occur, since THEY are prosecuting and not the plaintiff. Even if practically the malveh pursues the debt, he is acting as an extension or agent of beit din. The Rashbam in Bava Batra (65b-66a) cites Rashi's logic: "once he has transferred the collection of his debts to beit din... he no longer violates 'lo yigos'."
In fact, this mechanism of pruzbul might also be based upon a different halakhic precedent. We noted earlier that according to the Sifri, pruzbul appears to be based on the mashkon model of initiated partial collection. The gemara in Gittin (37a; also based on a Sifri) asserts that if someone ACTUALLY transfers his debts to beit din for collection, they remain unaffected by shemitta. The Sifri derives this principle from the term lo yigos - after transferring the debts to beit din the creditor is no longer the force behind the collection. Evidently, according to Rashi, Chazal modeled pruzbul upon this concept: by drafting a pruzbul, a person symbolically transfers the collection of his debts to beit din, thereby skirting the cancellation of those debts. In fact, the actual language of pruzbul cited by the mishna (Shvi'it 10:4) more closely resembles Rashi's definition of pruzbul then it does the Rash's. The mishna presents the language as follows: "I convey to ploni, and ploni, the judges of this court, that any debt of mine can be collected at any time." This language – particularly the word 'mosrani' (I convey) - implies the transfer of the collection of debts to the beit din.
Of course, this view of pruzbul does not directly account for the land condition stated above. According to the Rash, land allows the collection to be viewed as having already commenced (literally, 'ke-gavui dami'). According to Rashi, if pruzbul is merely the symbolic transfer of debt to beit din for collection, why should land ownership on the part of the debtor be necessary? Rashi in Gittin (37a) suggests a purely technical basis for the land requirement: most loans revolve around land as collateral. Only a minority of loans are tendered without land collateral. Since chazal generally do not issue decrees in unusual circumstances ('lo shechi'ach'), they did not apply pruzbul to a loan without land to back it up. The land requirement doesn't exhibit the essential nature of pruzbul; it rather reflects a purely technical factor relating to the general strategy of Chazal's takanot.
The Rashbam in Bava Batra develops a different concept. Indeed, the mechanism of pruzbul is based on transferring collection to the courts. From this standpoint, land should not be required. However, Chazal did not want to give the appearance of completely uprooting the effects of shemitta. Hence, they required a secondary condition to mitigate the 'effects' of their takana. As land simulates the mashkon effect of partial payment it was required alongside of pruzbul. The Rashbam agrees with Rashi that pruzbul works primarily through the conveyance of debts to beit din for collection. However, land owned by the loveh approximates the mashkon condition and 'moderates' the process of pruzbul.