This article claims to uncover the core problematics that have made the debate on defining and conceptualizing "religion" so difficult and argues that this makes it possible to move beyond radical deconstruction towards reconstructing the concept for scholarly purposes. The argument has four main steps. Step 1 consists of establishing the nature of the entity "religion" as a reified imaginative formation. Step 2 consists of identifying the basic dilemma with which scholars have been struggling: the fact that, on the one hand, definitions and conceptualizations do not seem to work unless they stay sufficiently close to commonly held prototypes, while yet, on the other hand, those prototypes are grounded in monotheistic, more specifically Christian, even more specifically Protestant, theological biases about "true" religion. The first line of argument leads to crypto-theological definitions and conceptualizations, the second to a radical deconstruction of the very concept of "religion." Step 3 resolves the dilemma by identifying an unexamined assumption, or problematic "blind spot," that the two lines of argument have in common: they both think that "religion" stands against "the secular." However, the historical record shows that these two defined themselves not just against one another but, simultaneously, against a third domain (referred to by such terms as "magic" or "superstition"). The structure is therefore not dualistic but triadic. Step 4 consists of replacing common assumptions about how "religion" emerged in the early modern period by an interpretation that explains not just its emergence but its logical necessity, at that time, for dealing with the crisis of comparison caused by colonialist expansion. "Religion" emerged as the tertium comparationis - or, in technically more precise language, the "pre-comparative tertium" - that enabled comparison between familiar (monotheist, Christian, Protestant) forms of belief and modes of worship and unfamiliar ones (associated with "pagan" superstition or magic). If we restore the term to its original function, this allows us to reconstruct "religion" as a scholarly concept that not just avoids but prevents any slippage back to Christian theology or ethnocentric bias. This article claims to uncover the core problematics that have made the debate on defining and conceptualizing "religion" so difficult and argues that this makes it possible to move beyond radical deconstruction towards reconstructing the concept for scholarly purposes. The argument has four main steps. Step 1 consists of establishing the nature of the entity "religion" as a reified imaginative formation. Step 2 consists of identifying the basic dilemma with which scholars have been struggling: the fact that, on the one hand, definitions and conceptualizations do not seem to work unless they stay sufficiently close to commonly held prototypes, while yet, on the other hand, those prototypes are grounded in monotheistic, more specifically Christian, even more specifically Protestant, theological biases about "true" religion. The first line of argument leads to crypto-theological definitions and conceptualizations, the second to a radical deconstruction of the very concept of "religion." Step 3 resolves the dilemma by identifying an unexamined assumption, or problematic "blind spot," that the two lines of argument have in common: they both think that "religion" stands against "the secular." However, the historical record shows that these two defined themselves not just against one another but, simultaneously, against a third domain (referred to by such terms as "magic" or "superstition"). The structure is therefore not dualistic but triadic. Step 4 consists of replacing common assumptions about how "religion" emerged in the early modern period by an interpretation that explains not just its emergence but its logical necessity, at that time, for dealing with the crisis of comparison caused by colonialist expansion. "Religion" emerged as the tertium comparationis - or, in technically more precise language, the "pre-comparative tertium" - that enabled comparison between familiar (monotheist, Christian, Protestant) forms of belief and modes of worship and unfamiliar ones (associated with "pagan" superstition or magic). If we restore the term to its original function, this allows us to reconstruct "religion" as a scholarly concept that not just avoids but prevents any slippage back to Christian theology or ethnocentric bias.