“I’m on the fence” means “I have not decided between the options.” Instead of treating indecision as unwillingness or agreement, ask which comparison still matters: “Are you weighing the cost, the time involved, or whether it suits your needs?” Then agree what information or deadline would allow a real decision.
The speaker may like both choices, dislike both, lack a key fact or simply need time. The phrase does not authorise someone else to choose for them. In ordinary English conversation, repair the ambiguity by naming the decision, identifying the unresolved criterion and confirming a reasonable next check rather than repeating “So is that a yes?”
Quick answer: a five-part decision conversation
- Identify the choice: “Are you deciding between the Saturday and Sunday sessions?”
- Find the sticking point: “Which part is hardest to compare?”
- Ask for the smallest useful fact: “Would knowing the travel time help?”
- Respect decision ownership: “It’s your choice; I can explain the options.”
- Set a truthful checkpoint: “We need a response by Thursday to arrange transport; is that enough time?”
The checkpoint should be a real constraint, not invented pressure. If a decision has no deadline, say so. If someone else will be affected, explain the practical consequence and ask for a choice without claiming that quietness means yes.
Did you know? A person on the fence is not literally sitting on a barrier
Cambridge Dictionary describes on the fence as being unable to decide something. Its image is a person between sides, but its conversational purpose is to report an unresolved preference or judgement. A literal translation can mislead an English learner who has never heard the expression.
The idiom can save face: rather than declaring that one idea is bad, the speaker acknowledges that judgement remains open. But it can also hide what information is missing. A respectful question changes vague indecision into a useful comparison without assuming that the speaker is lazy, indecisive by nature or secretly opposed.
Why the language works: position, reason, evidence, decision
“I’m on the fence” tells you a decision state, not the reason behind it. Listeners often fill the gap with a guess: “You must be nervous,” “You must prefer the cheaper option,” or “You don’t want to join us.” Those guesses are stories, not evidence. Ask one neutral question about criteria rather than inventing the reason.
Think of four separate layers. Position: a choice is not settled. Reason: one or more criteria are unresolved. Evidence: the chooser may need relevant facts, a demonstration or reflection. Decision: only the person or authorised group can confirm the outcome. Separating the layers keeps help from turning into pressure.
Fictional dialogue: two school project formats
Fictional teaching example. Jay: “I’m on the fence about doing a poster or a short video.” Lina: “Which part are you unsure about?” Jay: “A video might explain our Science investigation better, but we have little editing time.” Lina: “Then let’s compare what each format needs and check the teacher’s actual requirements before deciding.” Jay: “Good idea. We can make a quick outline for both.”
The group has not chosen either format. It has found a criterion—time and presentation requirements—that can be checked. The useful next step is comparison, not a vote based solely on who speaks first. Do not invent assessment rules; ask the relevant teacher or consult the real project instructions.
Fictional dialogue: a parent offers a choice without judging hesitation
Fictional teaching example. Child: “I’m on the fence about joining the art club.” Parent: “Is it because of the activity, the schedule or something else?” Child: “I like drawing, but I won’t know anyone.” Parent: “Would visiting once or speaking to the organiser help you decide?” Child: “I’d like to visit first.” Parent: “Then we can ask whether a trial visit is allowed. You can make your choice afterwards.”
A decision criterion may be emotional or social as well as numerical. The parent does not assume the concern is irrational or that every problem can be solved with a price comparison. The child names the support they need. If a trial is not available, the family can consider another way to learn about the group.
Fictional dialogue: friends decide where to meet
Fictional teaching example. Farah: “I’m still on the fence about the two venues.” Sam: “Is accessibility, cost or travel the main issue?” Farah: “The second is easier for my grandmother to reach.” Sam: “That matters. Let’s check whether both have suitable access before making the booking.” Farah: “Yes, and then I’ll confirm which works.”
This conversation is not solved by averaging preferences. A constraint affecting a participant’s ability to attend deserves investigation. A polite agreement to check facts is not yet a confirmed venue booking. Once the details are known, the person responsible for arranging the event should close the decision explicitly.
Fictional dialogue: an adult deciding whether to volunteer
Fictional teaching example. Organiser: “Would you like to help next month?” Volunteer: “I’m on the fence.” Organiser: “What would make the commitment easier to evaluate?” Volunteer: “I don’t know how many hours it takes.” Organiser: “The role needs two sessions of roughly two hours, plus the briefing. Would you like the full schedule?” Volunteer: “Yes. I’ll give you a firm answer by Friday.”
The organiser supplies a material requirement and a genuine deadline. They do not convert a request for information into a sign-up. If the person later declines, respect it. The conversational success is a better-informed choice, not a guaranteed yes.
Fictional dialogue: a team considers a software change
Fictional teaching example. Manager: “Are we ready to adopt the new tool?” Analyst: “I’m on the fence because we’ve only tested it with a small task.” Manager: “What needs to be demonstrated before you can recommend it?” Analyst: “Whether it handles our usual file sizes without losing comments.” Manager: “Let’s agree one test and review its results on Tuesday. That won’t automatically mean approval.”
Here the question seeks a testable criterion. A person may remain on the fence even after a test if further uncertainties matter. The decision owner must specify the standard and record the outcome; the idiom itself does not delegate authority to the most confident speaker.
When should you stop asking questions?
Not all hesitation needs to be overcome. The speaker may be deciding about something personal, or may already have stated that they do not want advice. If the choice is theirs and no deadline exists, a useful answer can be “That’s fine. You can decide in your own time.” Avoid asking the same question repeatedly in slightly different words.
There is also a difference between supporting someone to clarify a genuine choice and extracting a commitment from them. “What would help you compare?” invites thought. “If you really cared, you’d choose now” manipulates it. A clear outcome includes the possibility of a respectful no.
Tone and register: choose words to match the audience
- Casual: “Still on the fence? What’s making it tricky?”
- Gentle: “You don’t have to answer now. Is there anything you would like to know first?”
- School discussion: “Which requirement would help us choose fairly?”
- Professional: “Which unresolved criterion prevents a recommendation, and when can we review it?”
- Overbearing: “You’re avoiding the question; I’ll count you as agreeing.” This converts uncertainty into unauthorised consent.
Common misconceptions and conversation repairs
- “On the fence means no.” Repair: “I haven’t decided. Please don’t count me either way yet.”
- “On the fence means yes, eventually.” Repair: “I might choose either option; I need to compare them.”
- “There is always one missing fact.” Repair: “There may be several criteria or a personal value I’m weighing.”
- “The listener should solve the choice.” Repair: “You can help me see the options, but the decision is mine.”
- “Waiting is always harmless.” Repair: “If there is a real booking or safety deadline, state it and agree how to proceed if no answer arrives.”
Ask these questions before making a decision
- What are the real options—and is declining both an option?
- Who has authority to decide, and who will be affected?
- What criterion is unresolved: evidence, cost, effort, safety, values or something else?
- What fact or small experiment might genuinely change the choice?
- What is the real last responsible date for deciding?
- If the result remains undecided, what happens without pretending consent?
Retrieval practice: recover the missing part
Cover the examples and practise speaking. A classmate says, “I’m on the fence about which source to use.” A family member says, “I’m on the fence about travelling.” A colleague says, “I’m on the fence about approving the draft.” In each case, identify the missing criterion and ask one neutral question. Do not prescribe the answer.
Suggested responses: “Which source is more relevant and reliable for the specific question?” “Are you mainly weighing cost, accessibility or the dates?” “Which criterion or unresolved check prevents approval?” The exact criterion will emerge from the speaker’s reply; these are invitations, not assumptions about what the real reason must be.
A short exercise: two criteria, one checkpoint
Draw three columns headed Option A, Option B and Still unknown. Choose a harmless decision, such as reading a paper book or listening to an audiobook for a group assignment. Name two criteria, such as access and ability to review details. Ask what evidence would settle the unknown. End with one sentence that explicitly identifies who will decide and when.
Then transfer the method to a real conversation. When you hear “I’m not sure,” “I’m torn,” or “I need to think about it,” resist the urge to supply the missing reason yourself. The idiom is different, but the repair job is the same: identify the choice, the actual blocker and the next legitimate step.
Guidance for parents, students and teachers
Parents: make the child’s reasoning visible without turning every option into a test of obedience. Students: practise naming the criterion instead of saying only “I don’t know.” Teachers: model how a reasoned choice differs from a popularity vote or a confident guess. Ask students to describe which evidence or value changed a decision.
The broader distinction between learning and agreeing is explored in eduKate’s lesson on checking willingness without pressure. The “up in the air” lesson addresses a different job: a plan whose status is uncertain across participants, not necessarily one person weighing alternatives.
Frequently asked questions
Does “on the fence” mean someone is physically sitting on a fence? In this context, no; it is a figurative expression for being undecided.
Is being on the fence a refusal? No. It describes a current state of indecision, not a confirmed answer.
Can I ask why? Yes, but a specific neutral question about what matters is often more helpful than a repeated “Why can’t you choose?”
What if a decision is urgent? Name the real deadline, explain what happens without a response and seek an explicit choice or authorised fallback.
Does helping someone decide mean deciding for them? No. Clarifying evidence and options can support a decision while leaving authority with the right person.
Sources and further reading
Language: Cambridge Dictionary, “on the fence”. Conversation practice: British Council B1 speaking and its listening lessons. Conceptual foundation: eduKate What Is Vocabulary?. All dialogues here are invented teaching examples; they report no actual trial or measured outcome.