This is a case of an exceedingly common conundrum of insufficient specificity. That is, one can find support for contrary decisions given the same claim because a subset of cases may justify one decision while a different subset of cases may justify the opposite.
For example, consider the claim "Killing is bad".
- Refusing to kill a known serial killer lets him roam free and
continue killing. But killing is bad.
- Killing a known serial killer
(capital punishment) results in fewer innocent deaths, and no
innocent deaths are incurred by the murderer's executor. This means that killing the perpetrator of killings of the innocent is good in terms of minimizing killing--a seeming contradiction owing to different qualifiers on "killing" being compared as though it were apples to apples, but it's not.
Therefore the claim "killing is bad" finds both support and rejection depending on the situation. It's related to Simpson's Paradox, in which the decision we make depends on the subset of the data that we are looking at. Therefore, preserving this context as a qualifier on the claim is a crucial piece of data that enables non-contradictory resolution. Otherwise, the unqualified statement is over-generalized, finding both correct and incorrect applications.
Armed with the distinguishing features between cases that support or reject our premise, we can refine the premise. In this case, we can intelligently and correctly conclude:
- Killing innocent people is bad.
and
- Stopping killers of the innocent is good.
In terms of overall good, the taking of the life of a murderer is justified because (a) he is not innocent, and (b) it can decrease the overall amount of innocent death.
The premise you presented can be refined in part using the conditions already supplied in the justifications:
- Abortion is either right or wrong. (Given the appropriate qualifiers, yes).
- If it is right, then people should be allowed to kill fetuses.
- Killing fetuses is wrong.
- Hence, abortion is not right.
If it is wrong, then women won’t have the right to decide if a fetus
should grow inside of her.
This is the false premise. She can have the right to decide if a fetus should grow inside of her, independent of whether elective abortion is right or wrong. Consensual sex is consent to allow a fetus to grow inside the woman's uterus should fertilization occur. Once this decision is made, it cannot be revisited without harming either the mother or the child. It is commonly acknowledged that the irreversibility argument applies at least to children who have already been born and there would be no legitimacy in claiming a retroactive effect, denying the child's right to life.
Women do have the right to decide if she wants to develop a fetus
inside of her.
True. See my response to #5.
Hence, abortion is not wrong.
The incorrect conclusion stems from premise 5. She has the right to decide whether to host offspring, independent of the judgment of whether or not she can subsequently reverse it through abortion.
We can see that morally and philosophically, ethically and legally, it is possible to enter into a legally binding contract but not to discard it without a sufficient justification. Therefore the avoidance of the conclusion that consensual sex is consent to become pregnant, as well as the claim that a woman can reverse the former agreement without a sufficient justification (e.g., her life being seriously threatened by the pregnancy) are invalid. (Related: viable, safe pregnancies are frequently misdiagnosed as inviable, unsafe ones)
So the conclusion
C. --Abortion is both right and wrong.--
Should be enhanced to
C. Elective abortion of mutually viable life is wrong. (Elective abortion means a reversal of a prior conscious decision by the mother to risk conception, to the extent of exterminating life she had begotten by that willing act. Mutually viable means that both mother and child are accurately considered as having a good chance of survival.)
This leaves the remainder:
- What about cases in which the mother became pregnant by nonconsensual
sexual activity, including if she was a minor?
We can apply the same procedure to this question as to the former, by discovering whether any condition exists that mitigates or reverses the conclusion held to in other cases. Women who became mothers without their own consent to sex can in many cases still bear those children safely, and charitable help is available in many areas for such. In such cases, adoption agencies can also help them if they are unable or do not wish to bear the burden of support for the child.
- What about cases in which the mother's life is indeed threatened by the pregnancy?
Engage in the same pattern of value consideration as above. Intelligent, compassionate souls often find ingenious ways to preserve the good, and life often finds ways to thrive even in seemingly hostile environments and against incredible odds. One of the very difficult problems of life is that we cannot guarantee everyone's safety or immunity to illness, disease, and death. Charitable help and respect for those making very difficult decisions is needed here.
- What if the life of the child is not viable?
If the child is merely deformed but cannot live, respect for life demands that we still try to give the child every opportunity and comfort that is within our power to give. Many such efforts have been rewarded with miraculous success. Some have not, or at least not so immediately.
Does this mean that it is possible for the premises to be true and the conclusion false for this reason?
One of the premises above (number 5) has been shown to be false; it is an example of too narrow a scope, or erroneously presuming that the proposal is the sole solution to the query.
I only numbered it to help see the premises and conclusion. I'm asking if this argument as a whole is valid or invalid in terms of philosophy.
The argument as a whole is not valid. It suffers from a very common logical fallacy of failing to consider alternative means to satisfy the stated conditions. In Bayesian terms, we might call this failing to normalize by the marginal--that is, the significance of one claim or possibility is overstated because alternative possibilities or claims are excluded from consideration.
A litany of other interesting examples can be examined, having this property:
- "Equality is good" -> Equality of what?
- "Diversity is good" -> Diversity of what? Goodness and badness, or only of goodness?
- "Pain is bad" -> ...
And so on.