As for dictionary entries: the Cambridge English Dictionary defines “secular” as “not having any connection with religion.” In the Oxford English Dictionary, it is “not connected with spiritual or religious matters.” In Macmillan, “not religious or not connected with religion.” In Collins, “things that have no connection with religion,” and many further examples could be cited. In ordinary usage, therefore, a “secular religion” would mean a “religion that has no connection with religion” or a “not religious religion.” The vague reference to “spiritual or religious matters” is worthy of more profound investigation (into the difference between “spiritual”, “religious”, “transcendent”, “supranatural”, and other similar words), but I will return to this issue later.
A curious exception is Merriam-Webster’s Dictionary in which “secular” is “not overtly or specifically religious.” By a closer look, this definition, which seems only a weaker version of the ones just cited, in fact contains a very strong statement (if we take it at face value): that what is called “secular” is not necessarily non-religious, only covertly or non-specifically so. Secular religion theorists would no doubt find this approach appealing, but the same dictionary also says that “secular” is “of or relating to the worldly or temporal”, which once again leads to an inconsistency, because “worldly” – according to the same dictionary – is “of, relating to, or devoted to this worlds and its pursuits rather than to religion or spiritual affairs.” So, despite all attempts to find more and more complex definitions, secular religion theorists are still left with the notion of a religion which is devoted to other things than religion (or “spiritual affairs”, whatever the latter means).
At least if they rely on dictionary definitions. In academic literature, the definitions of “secular” put more emphasis on the historical use of the word, which seems to offer a way out of the contradiction. Since “secular” comes from the Latin saeculum, “century” or “age”, the longest time span of a human life (Holford-Strevens 2013), its original meaning does not necessarily exclude all reference to things that are nowadays called religious. In this case, secular would mean something “temporal”, so the contrast is not between the secular and the religious, but between what is connected to time and what is not. Thus (unless one insists on equating the latter with the religious), speaking of “secular religion” would not be a contradiction in terms, only a specific form of religion, provided that ancient authors made such a distinction between temporal and non-temporal religions. The fact is, of course, that they did not, and when a Christian author like Augustine of Hippo used the Latin words of saeculum and saecularis, he did it exactly to separate worldly or terrestrial things (the city of man) from the city of God, something that most of us today would associate with the realm of religion (Beaumont–Eder–Mendieta 2018, 3).
So – even though Augustine did not have the same concept of religion as we do today – the contradiction remains there. The fact that the Middle Ages again used the terms “secular” and “religious” in a different sense will not help to eliminate the contradiction, either. Every medievalist knows that saecularis meant the opposite of “ecclesiastical” in general, or “monastic” (also called religiosus) in a more specific sense (Niermeyer 1976, 951). A secular religion – if there had existed such a word – would therefore have meant something like a non-ecclesiastical religion, which was unimaginable in the medieval context, or once again a non-religious religion which would have been an obvious logical contradiction.
Let me emphasize once more that I am not occupied here with the constantly changing definitions of words like “secular” and “religion” in themselves, but with their relation to each other. What is true today seems to have been the same throughout history: “secular” was always an opposite of “religious” and moving beyond the oxymoron would only be possible if we took “secular” in one historical meaning, and “religion” in another. Apart from the terminological confusion that such an approach would cause, it is also true that no one writing on secular religions relies on such dubious maneuvering. Regardless of how medieval and modern texts define the secular and the religious, secular means non-religious, and religious means non-secular in both contexts.
In a similar way, no academic literature on secularization would be possible without defining the secular in opposition to religious consciousness, behavior, and institutions (Wilson 1982, 149); as not connected to religion as “faith in, or adherence to God, or some notion of ultimate reality” (Taylor 2007, 1); as an alternative to religion (Bruce 2009, 146); or as “a realm or reality differentiated from the religious” (Casanova 2011, 54); but this is perhaps already overstating the obvious. Whether one takes dictionary definitions, historical accounts, or analytical classifications as a starting point, it remains true that “secular religion” is an oxymoron, the worst possible choice when describing self-professedly non-religious ideas, actions, and institutions as being similar to ones that are more commonly – albeit vaguely – called religious.