Normally, the future determiner form of a verb has a ㄹ directly attached to the base if it ends in a vowel. When this is not the case as in 저을, you can infer that the dictionary form of the verb has a root ending in ㅅ. For 저을, it is 젓다.
I paper I read a while ago suggested that these syllable-final ㅅ became ㅿ. Other papers I’ve come across show that syllables that had historically used ㅿ replaced it with ㅅ, ㅈ, or ㅇ in the modern language. I would be grateful if anyone could find that first paper again as I was only able to view it in Google’s cache and the web has a tendency to disappear things from existence never to be found again.
EDIT: I found this: https://m.blog.naver.com/kiwimino/220735250288
④ 젓다: 자음으로 시작하는 어미 앞에서 ‘젓-’으로 활용 (예) 젓고 모음으로 시작하는 어미 앞에서 ‘저ᇫ-’으로 활용 (예) 저ᅀᅥ → 여기서 ‘ㅿ’은 ‘세모’가 아니라 ‘반치음’으로 ‘z' 소리이다. 그런데 'ㅿ’ 역시 제한된 환경에서만 사용되다가 16세기 들어 사라지기 시작한다. 그로 인해 ‘저ᅀᅥ’에서 ‘ㅿ’이 탈락해버려 ‘저어’가 되어버린다. ‘젓다’→‘저어’, ‘짓다’→‘지어’와 같은 현대국어의 ‘ㅅ’ 불규칙 현상은 바로 이와 같은 역사적 변화에 따라 형성된 것이다.
Long after the disappearance of ㅿ, its invisible presence is still felt as a contraction blocker.
This paper aims to demonstrate that the intervocalic ‘s>z’ is a lexically gradual sound change and the /s/-preserving dialectal forms are the residual forms of the change. In this paper, I approached the ‘s>z’ change from the perspective of a gradual lexical diffusion. I showed the ‘s>z’ change has diffused in the lexically gradual way over a long time span drawing on the investigation of various dialectal forms related to /z/(=ㅿ) of Middle Korean. I especially considered the ø-forms and /?/-forms of Contemporary Korean dialects which were originally /s/-forms in Middle Korean as the important evidence of the ‘s>z’ change, since they are the results of gradual sound changes such as ‘s>z>ø’ and ‘s>z>?’. Furthermore, I approached the question why ‘s>z’ change ceased from the phonetic viewpoint. I suggested the gradual increasing of the degree of aspiration of /s/ caused the decreasing of the possibility of intervocalic voicing of /s/.