Most importantly, please read the article by Hyams and then the section by Hudson.
The principal issues involved with warranty are:
Warranty is crucial to the understanding of the personal and social relationships behind the tenure of land in the 12th century which became formalised as legal obligations. Milsom's perception, it seems (though one is never entirely sure), is that warranty concerned the relationship between lord and tenant, a personal obligation from the lord to the tenant. First, provided the services were performed by the tenant, the lord who put the tenant in (seised) had an obligation to protect the tenant, and, secondly, provided those services were performed, the lord forfeited any right to put out (disseise) the tenant. Milsom's principal example of warranty is then: the lord A enfeoffs B with the prospect of succession by B's heirs; but B disappears and it looks like the tenure is at an end, so lord A enfeoffs C; B's heir then appears and requests admission to the tenure. Lord A must then provide equal compensation for either B or C, the escambium (exchange), so that warranty and exchange are inextricably intertwined.
Milsom's important interpretation was only partly accepted by Hyams, who searches for a different explanation of the origins of warranty. Whilst accepting that warranty became associated with tenurial relationships, gifts/grants of land, and 'good lordship' -- related to the duty of lords to provide maintenance and protection -- Hyams prefers an alternative explanation of how that association developed. Warranty existed before it was subsumed into tenurial relationships and a guarantee of land; its origins seem to have been in Anglo-Saxon chattel warranty (at least as early as the 10th century), so that warranty as such originally guaranteed that the chattels sold were owned by the seller. The primary principle behind warranty was thus that it was the owner's to give and warrant. That principle became transferred to tenurial relationships in land.
Hyams perceived that two aspects came to inform warranty: the 'negative' protection afforded by warranty, that is, against the lord and his heirs -- that the land was the lord's to give and that neither he nor his heirs would infringe the tenant's right (provided the services were duly performed) -- that is, no resumption without due cause; and the 'positive' defence provided against third parties -- the rest of the world -- contra omnes homines. Exchange thus became an inherent part of the positive aspect of warranty -- the lord's obligation to provide an exchange of equal value should he fail to protect the tenant against outside claims to the land 'an automatic corollary to warranty' (Hyams, p. 465).
Hyams assumed that warranty was ab initio an hereditary obligation binding on the heirs of lords, but his argument was largely by analogy to the presumed implications of warranty of free alms. Warranty of free alms 'showed everyone that warranty could be a continuing commitment' (Hyams, p. 472). [That supposition may be subject to some other considerations: (i) warranty in this case may essentially be warranting the exception from secular services, since many early warranty clauses involved with free alms seem to be connected to the services; (ii) we should be cautious of inferring the nature of secular relationships from gifts to religious houses which may have been considered exceptional before 'Glanvill' defined alms as simply another form of service].
Hyams thus concludes, importantly, that warranty and exchange came to equate with tenant right:
Hudson confirms much of the suggestions of Hyams, in particular accepting that the origins of warranty lay in the sale of chattels, so that the primary function of warranty between lord and tenant was related to the notion of sale of chattels -- that the land was the lord's to give. Hudson refines the argument of Hyams by implying that exchange was not necessarily an integral part of warranty at its inception, and here he differs over the interpretation of the clauses in charters of St Mary's, York, which exclude exchange without mentioning warranty as such. He also places more emphasis on the positive force of warranty as effective against the rest of the world.
In addition, Hudson refines the chronology of the development of warranty in charters, as Hyams largely considered the collection of charters in Early Yorkshire Charters. The development of warranty is an important consideration, of which the following are the principal points:
Hyams reached the conclusion, however, that warranty as an explicit clause in charters was not frequent before the 1170s and still not completely the norm at the end of the 12th century. He proposed that the more expansive use of warranty clauses in charters was propagated by religious houses. Although they did not recognise lordship (accordingly with the canons of the Church), yet they adopted any practice for the confirmation and security of lands given to them, including warranty clauses.
The nuance which must be placed on this last thesis is the appearance of the warranty clause in some charters of the 1140s, mainly of the Earl of Chester, and 1150s, including other lords, which intimate that, at this time at least, warranty was particulary associated with 'good lordship' at a time of maximum insecurity of tenures of mesne tenants -- the 'Anarchy' of Stephen's reign and the prospect of disseisin through the territorial disputes of magnates.
Warranty is thus of great consequence for the development of tenant right against discretionary lordship, although how it evolved is a matter of different interpretations.