This corpus-based study aims to identify the interactional and interactive metadiscourse markers in terms of frequency in the abstract and discussion sections of research articles on linguistics, written in English by native, Egyptian, and Saudi researchers. To attain this aim, 60 research articles have been randomly compiled and analyzed qualitatively and quantitatively via AntConc.3.2.4 depending on Hyland's (2005) classification of metadiscourse markers (MM). Taking the abstracts and discussions written by the natives as a benchmark, this study poses the following essential question: How close and far is the amount of the interactional and interactive resources in Egyptian and Saudi abstracts and discussions to and from the native level? The results showed that except for hedges, evidential markers, and endophorics, the usage of attitudes, code glosses, engagement markers, self-mentions and transitions in the E-abstracts (i.e. written by Egyptian researchers) was much far from the native level. But in S-abstracts (i.e. abstracts written by Saudi researchers), only two close points to the native level have been recorded: transitions and engagements. In the E-discussion sections, unlike code glosses and frame markers, attitudes, boosters, endophorics, hedges, and self-mentions were reported very close to the N-level. In the S-discussion sections, boosters, code glosses, emphatic, engagement, frame markers, and transitions have recorded far rates from the N-level; whereas only attitudes and hedges were much close to the native normal level.