A recent question got me thinking about how to criticize poetry well and I realized I am not very good at it. I can do themes, and that's about it. And I'm not even sure that should be covered unless explicitly asked for as was part of the issue in the linked question. So what do you look for? Do you talk about meter and rhyme schemes? Do you mention their use of slant rhyme isn't as successful as they intended it to be? How do you make it a scholarly activity that leaves personal preference for things like uplifting tone completely out of the equation?
I studied literature at school, and the most important thing I learnt is that the obejctive of a poet is to transmit something, whether an idea or a feeling. To do so, the poet uses a range of techniques to craft their poem.
While reading, one can experience the idea or feeling in a leisurely way, at an instinctive level, or one can dive into the poem and consciously understand how the whole works. It's a bit like cars. I don't need to understand the mechanics to enjoy the sound of an engine, but my father's enjoyment of a car would never be complete unless he understood the mechanics that produce those particular sounds.
So, if you want to critique a poem in a structural way, here's my suggestion:
read it and feel it (which means you can enjoy it or dislike it)
identify the topic and the theme
break the poem into its parts (stanzas or lines) and see how the theme evolves through the poem (eg.: it can start with a chirpy happiness that dwindles into melancholy)
look at the rhyme and the rhythm (it helps to know technicalities such as rhyme schemes and metre) and see how they underline or oppose the ideas
look for rhetoric figures and see how they underline or oppose the ideas being transmitted
If you don't know any rhetoric figures, focus on the following points:
a) parallel and opposing structures (in sentence structure, but also in sounds, ideas, ...)
b) repetitions and contrasts (of sounds, adjectives, ideas, ...)
c) comparisons and metaphors
d) general images and allegories
e) connotation of words
Remember that the important thing is the effect those figures transmit.
Now, repeat the first step: read and feel the poem. Do you still get the same feeling of the first reading? If so, great. If not, what changed?
Finally, decide why you like it. If you happen to dislike it, it is even more important to identify the reason:
Is it about the general message?
I typically dislike 'in your face' messages, as I prefer subtlety in poetry, with images, allegories and symbolism. If this is the case, feel free to explain your lack of enjoyment as being caused by your personal likes rather than the poem itself.
Is it a particular technique or structure you dislike?
I don't typically enjoy poems with long stanzas and I really hate poems that go for a disharmonic vibe. It's a tool like any other, but I can't stand it. If this is the case, again feel free to explain the reason for your lack of enjoyment. Remember that it's hard to analyse how well the author used their chosen techniques if you feel a strong dislike for the work, but make sure you tell them it's not a problem with them or their work, it's really about your tastes getting in the way of impartiality.
If it's none of the above, did the author fail to use the best tools (or used them badly), giving you a conflicting feel of the general work?
This may require you to repeat the analysis of the whole piece looking for what sparks the dislike. Once you identify it, check one last time if the problem was poor skill on the author's part or if you simply don't enjoy that particular technique.
If you are still confident the problem lays with poor skill, point out how the words (or lines, rhymes, ...) create an effect that doesn't help transmit the message.
As an alternative to the long process described above (although, with practice, you can do this in about 5 minutes for a general analysis and 15-30 minutes for an in-depth one), you can ask 'how does the poem create feeling X?'. Of course, you'll still end up covering most of the points.
Remember that very word - and every missing word - means something, whether the author meant it or not. Some examples:
If the lines are short, they can either give a feeling of lightness or of brevity.
If there are a lot of short, open vowels (eg.: cat, sit), then the tone is likely to be lighter, but if the vowels are long and closed (eg.: door, loose) then the tone is likely to be heavier.
Wide waves wandering
in the dark sea.
In this short example, the first line has the 'w' sound repeated at the beginning of each word with a rhythm (in sound and sight) that reminds one of the shape of a wave. Being a long line also underlines the spacial idea of 'wide' (because the poem is wide).
The second line is short, being a single word, and gives a sharp contrast to the other two, which are longer. It makes that idea more powerful. Notice also the vocalic sounds at the beginning of the words in both lines and how that similaritude connects the two lines in the absence of rhyme.
The words 'aimless' and 'dark' have a negative connotation here. 'Wandering' becomes negative, too, especially because 'wandering' and 'aimless' work together in a loop which the word 'waves', underlined by the repetitive 'w', helps to conjure into the readers' mind.
Please note that traditional word connotations can be undone by the context. The word 'dark' is often assumed as negative and, yes, it can be negative when talking about death, but it can also be positive when talking about night skies or a lover.
Finally, if you find that the poem you're criticising is about death but seems to have a multitude of positive words and ideas while the lines and the rhythm give you a light-hearted feeling, perhaps point it out and ask if that was the intention because, to you, it felt like the tone was detracting from the general effect.
I have my own three-part rubric I've used for years for critiquing any creative project:
Craft: How skillful is it? What command of the technical basics does the creator have? In terms of poetry, if it is a formal poem, does it meet all the rules? If the rules are broken, are they broken for good reason? Is it presented well, and in a way that is free of mistakes?
Creativity: Is it original? Is it fresh? Does it break new ground? Does it take risks? Is it memorable? Is it surprising?
Depth: Does it move the audience? Does it take on challenging subject material? Does it have deeper levels and layers of meaning? Is it heartfelt? Is it important or significant?
Using this rubric, there's almost always something to praise AND some valuable avenues for improvement. It's equally applicable to all art forms, and it works towards well-rounded artworks that excel across a wide variety of criteria. I've also found it to be successful in guiding people who wish to improve their art, which should be the end goal of all criticism. Remember, however, it's just as important, if not more so, to find the things that are being done well as it is to identify the areas for growth.
While critical analysis could get into a lot of factors - structure, vocabulary, form, relations to events and culture, use of poetic language forms, rhythm, flow, themes, all these things that comprise a poem, they are all secondary, and a simple, crude, primitively written couplet can sometimes have a much stronger impact than the most polished, fancy and advanced sonnet or epos. And this is all poetry is about - evoking specific feelings, causing a specific emotional impact - everything it does serves that singular purpose. If it achieves the effect, it's good. If it fails, or the effect is underwhelming or different than intended - say, it bores you, or gives a bad aftertaste of emotional manipulation - then it's bad.
Your critique may include analysis, how the effect is achieved, or what factors prevent the poem from creating the intended impact, and such analysis can include all the elements you want to include, but you must at all times remember they are subservient to the primary goal, just means, not ends.
How do you make it a scholarly activity that leaves personal preference for things like uplifting tone completely out of the equation?
Just like you respect an opponent in sports playing better than you; a rival in a debate pointing out your errors, or trapping you by clever use of eristics. It's wrong to leave out your feelings, because poetry is all about them - but of course the intended impact may be something you personally dislike - e.g. the poem being a pointed satire on a subject you hold in high esteem. In that case treat the poem as a worthy rival, underline strong points, show the weak ones, acknowledge the impact - you don't have to like it or agree with it, but it still works as intended!
A poem [implicitly] defines the criteria of its own success
This is actually largely true of any piece of writing. If you were to measure Kafka's The Metamorphosis and Frankl's Man's Search for Meaning by the same yardstick, you're not really being reasonable. A piece absurdist Magical Realism about the pointlessness of a particular life, and a holocaust survivor's rumination on finding meaning in spite of incredible suffering and tragedy, are interesting to contrast, but are intended to have quite different effects. The first question in either case is whether they succeeded. By objective measures, both did. Full disclosure: I hate Kafka's writing.
More specifically speaking of poems, you can't really critique a piece of free verse for not observing iambic pentameter. (You might, on the other hand, suggest that a particular long-lined free verse poem just sounds like pretentious sentences, and doesn't really resemble your understanding of poetry.) On the other hand, if something is a Shakespearean sonnet, except for garbled meter in a couple of the lines, then that metrical failure is a failure indeed.
Except in the subset of cases where the poet is carefully and intentionally setting up and then breaking expectations, it is the patterns and conventions established by the poet which should your first yardstick.
So yes, a remark that, for example, "Days" and "Skies" is not an adequate slant rhyme may be a fair criticism... Or it might not, depending on how rigorously the poet tends to rhyme, and how out-of-place the failure is.
This sliding-scale for criticism is especially important in regards to poetry, which can layer in any number of extra dimensions beyond what is usually put forward in prose, from meter and rhymes to specific families of images, or kinds of veiled commentary. Success is success of intention - when intention can be discerned.
There is a [fuzzy] line between quality and taste
In an essay on literary criticism, C.S. Lewis remarked that reviewers must be very careful in talking about a genre or subject they simply don't care for, and maybe should altogether avoid giving criticism on topics outside of their interest. There's a difference between "I don't care for spy thrillers," and "This was a bad spy thriller."
I could not fairly review Kafka; I don't understand how anyone could think anything positive about The Metamorphosis or A Hunger Artist. But there are people who think well of these pieces (at least well enough to assign them as reading in a that lit class I was taking years ago). I know, intellectually, that there must be some kind of dimension that other people experience beyond the vacuity and navel-gazing pointlessness that is all that I can observe in them.
Personally, when there's nothing about the theme that I feel that I can fairly critique, I do my best to focus on technical issues, or on the authorial intention, if I can manage it - but...
Ideally, one could state when differences in taste overwhelm objective measures
Poetry is a tricky thing.
It works purely on the level of language, by stretching its power beyond the simple "normal" direct communication (like you would have in prose). A poem does not just "say things", nor it just say things "in rhymes". A poem works with different many levels of language that it's hard to understand and to judge easily, like you point out.
Some elements are intrinsic in the poem, so they are easy to identify: metrics, rhythm (remembering that no classic metrics is still metrics), sounds (assonances, rhymes, and such), phrasing (nominal style vs verbal style, etc.). The second level goes to the words chosen and the images they evoke. Some poems use plain language and usual metaphors, other work by juxtaposition of different concepts and images, up to the abstract and symbolic plane.
Ultimately, it all goes to a deeper and invisible layer which requires a lot of culture and understanding of the context. To understand a work of poetry you need to understand the cultural context where it's born. There is a reason why Eliot writes differently than Shakespeare, or Zanzotto writes differently than Foscolo. It's not just about style, or age, but there is a poetical theory behind. Also, if you write a poem today in the same style and language of any masterpiece of 19th century, it won't make a good poem, it will be just ridicolous.
There is a lot beyond the form, and all of this you cannot just tell by reading the poem.
I know it sounds very much anti-romantic, but it's too naive to believe that a poem (as any work of art) lives in itself. I also love to abandon myself into the words and images, but it's too much a simplistic solution. Being a complex and synthetic form of expression, it requires some external hints to fully grasp the meaning and the quality of it.
+1 SF, emotional impact is important.
But basically the advice is the same: Stay technical. If on first impression, you "don't like" a poem, ask yourself if you weren't intended to like it. To take an extreme fictional example, if the poem is celebrating the assassination of a hated politician, I'm not going to like it. But I can realize I don't like the poem because of its meaning.
To do a proper critique, we must get over that personal emotional reaction to the topic, and critique the technicalities of poem formation intended to create that reaction.
I have not studied poetry, so you may have more examples than I; but I have read rhyming poetry where the rhymes are strained; or the metre is messed up, or words seem forced into the line -- They may be important but they don't fit. The same thing with syllabic elisions (using an apostrophe to replace a voiced letter or eliminate a syllable for the metre or count -- ne'er for "never", am'rous for "amorous", 'tis for "it is"). To me, these can be overdone and seem like crow-barring a word into a poem.
I know there can be patterns emphasis, and patterns of imagery. I know there can be echoes of imagery; e.g. a progression of flower names intended to symbolize various things; purity, virginity, love, death. I know a poem can feel rushed or feel crowded, a result of trying to pack too much story into the wrong form. Poems can be just as much a victim of cliché as novels.
As SF says in their answer, the emotion evoked in the reader is important, but regardless of the emotion, what are the mechanics of evoking it?
Just like when writing a novel, I don't ever want my reader to break their reverie of reading because their attention was drawn to bad writing or mistakes. I want to sustain that flow. I should think in poetry your aim is the same: If the whole point is to manipulate the reader's emotions, then any stray emotions, like irritation at a bumpy ride in the metre or rhyme, is a failure of the poem, because unintended emotion intrudes on their reverie.
That is what you are looking for in critiquing poetry. You want to be led through their their sequence of imagery and emotions and even surprises smoothly, and what you critique is anything that makes the trip not progress smoothly, that seems out of place, that stalls or grates or interrupts the reading. And, where their are opportunities where the author could apply some imagination to make a line or image have a larger impact.