© 2014 Elsevier B.V. Many, and perhaps all, oceanic island basalts (OIB) clearly contain a component of crustal materials that have been returned to the mantle through subduction or other processes. One of the first recycled materials to be identified as a potential source of OIB was mid-ocean ridge basalt (MORB), and this was later fine-tuned as having a long time-integrated (b.y.) high U/Pb ratio or high μ (HIMU) and producing OIB with the most radiogenic Pb isotopic ratios (206Pb/204Pb>20). However, it is becoming more evident that the compositional connection between subducted MORB and HIMU basalts is problematic. As an alternative hypothesis, a small amount (a few %) of recycled Archaean marine carbonates (primarily CaCO3) is proposed to be the main source of the distinct 206Pb/204Pb, 207Pb/204Pb and 87Sr/86Sr isotopic and major-trace element compositions of classic HIMU and post-Archaean marine carbonates for younger HIMU or the so-called FOZO mantle source. As an extension of the hypothesis, a conceptual model that combines the separate evolutionary histories of ancient oceanic lithosphere, which is the source of OIB, and upper mantle, which is the source of MORB, is also proposed. The model claims that FOZO mainly consists of the lithospheric mantle portion of the ancient metamorphosed oceanic slabs that have accumulated in the deep mantle. Such an ultramafic source is geochemically depleted due to prior extraction of basaltic melt plus removal of the enriched subduction component from the slab through dehydration and metamorphic processes. Combined with other proposed models in the literature, the conceptual model can provide reasonable solutions for the 208Pb/204Pb, 143Nd/144Nd, 176Hf/177Hf, and 3He/4He isotopic paradoxes or complexities of oceanic lavas. Although these simultaneous solutions for individual paradoxes are qualitative and non-unique, these are unified under a single, marine carbonate recycling hypothesis.