Reading Poetry

Share this essay on:

Alan Jay Lerner, the playwright-lyricist of Gigi, Camelot, and My Fair Lady, wrote a delightful autobiography, The Street Where I Live, which is filled with anecdotes.  One is relevant here.  Irving Berlin was trying to write a finale for a review, and he rushed excitedly to Moss Hart’s apartment to say that he had found one.  A self-taught pianist, he sat down at the piano and played his new song, “Easter Parade.”  It sounded terrible, and Hart didn’t know whether it was any good or not.  Finally, he saw a way out of his dilemma and asked Berlin to play one of his greatest hits, “Blue Skies.”  That sounded terrible, too, so Hart felt he could say:  “Irving, the finale is terrific.”

This story is very revealing.  It explains, at least in part, why Robert Penn Warren, Ezra Pound, and others are such terrible readers of their own poetry:  poetry reading is a performance art, and the original creators are not necessarily good performers.  One poet and critic—mercifully, I forget whom—actually said that when he read poetry he “put on his Deacon’s voice.”  What a ghastly idea!

But lack of performing talent isn’t the whole problem.  Most people know too little about poetry to read it effectively.  When you see one of Shakespeare’s plays, for example, you can only expect two or three of the actors to speak their lines really well, and all of them are performers.  This is not surprising.  Since about 1900 poetry has seen hard times.  Almost everybody writes prose libre (free “verse”) but hardly anyone writes traditional poetry.  Having become unfamiliar to people’s ears, it is difficult both to read and to understand.  This inability is little less than a cultural tragedy for it separates us from 600 years of our literary history.

The second essential in reading a poem is to understand what it says, and to use that knowledge to shape the phrasing and the emphasis.  But the first is to know that poetry is measured out by the line, and that if we are not aware of the lines it begins to function like prose.  Let’s look at some examples of iambic pentameter, the most common English line.  An iamb is a combination of a weakly stressed and a more strongly stressed syllable [dih DAH], so an iambic pentameter [penta-five, meter-measure] line is measured out in five iambic feet and has the rhythmic pattern:  [dih DAH] [dih DAH] [dih DAH] [dih DAH] [dih DAH].  The lines in the example given below are grouped into a stanza called a heroic couplet, a pair of rhyming pentameter lines which expresses a complete thought.  (A stanza is a set pattern of lines that make up part of a poem, or which may be a poem in itself.)  These come from “Mac Flecknoe,” John Dryden’s brilliant satire upon his rival, Thomas Shadwell, and they are spoken by the Prince of Dullness who is imagined to be looking about for an appropriate successor.

Shadwell alone my perfect image bears,

Mature in dullness from his tender years;


This is wonderful stuff.  Let’s see how the underlying iambic pattern may be varied in the actual reading.  The iambs (and substitutions for them) are enclosed in brackets.

[SHAD-well] [A-LONE] [my PER] [-fect IM] [-age BEARS],

[ma-TURE] [in DULL] [-ness FROM] [his TEN] [-der YEARS];

His name, “Shadwell,” requires a strong stress on the first syllable.  Both syllables of “alone” are stressed—there is no one like him:  he stands A-LONE.  The comma shows that the sense continues beyond the first line so we don’t want to end it with a falling intonation.  The second line ends in a semicolon—a much more definite pause—and here it would be appropriate to read the last syllable with a falling intonation.  Aside from all of this, we must slightly prolong the last syllable of each line.  This gives the rhyme more effect, and marks the line-ends, allowing the lines to work as units.  We must do this at the end of every line in every poem, whether the last syllable is a rhyme or not.

There is nothing arcane about any of this.  We would do the same sort of thing if we were trying to bring out all of the meaning of a prose speech, but every choice we make here acquires a special significance because it stands out as a variation on the underlying pattern.  Nothing of the kind can be said of prose (including prose libre) since there is no pattern to highlight the changes.

Shadwell, my spitting image, was precociously dull from the time he was small;

I’ve tried to preserve the meaning as much as possible, but a great part of the effect actually stems from the poetic form, itself, and can’t be paraphrased in prose at all.  Prose, on the other hand, can be easily paraphrased.  I might have said:

When Shadwell, my alter ego, was small he showed a marked talent for dullness;


Shadwell’s just like me, and when young you would have thought him fully mature in dullness.

These are virtually interchangeable as prose, but they can’t replace the couplet.  Why not?  The couplet paces the lines—setting up the blow and then delivering it.  This comes partly from the rhyme.  To see that this is so, let’s put the couplet into blank verse (unrhymed iambic pentameter):

Shadwell alone my perfect image bears,

Mature in dullness since he was a tot;


The pacing is there, but not the punch.  We need the rhyme to drive home the point and to connect everything into a tidy package:

Shadwell alone my perfect image bears,

Mature in dullness from his tender years;


Gotcha!  And notice that the lines do need to be the same length:

Shadwell alone my perfect image bears,

Mature in dullness from his very tender years;


This simply doesn’t work.  What we’ve discovered is that the only way to paraphrase this couplet and preserve its effect would be to write another couplet, and as it happens, that is exactly what Dryden does:

Shadwell alone of all my sons is he

Who stands confirmed in full stupidity.


Heroic couplets provided an easy starting point for us since many of the lines can stand by themselves.  Let’s look at a more difficult form in which the lines are enjambed, which means that the sense runs on from one line to the next.  The poem that comes to mind is Browning’s “My Last Duchess.”

That’s my last Duchess painted on the wall,

Looking as if she were alive.  I call

That piece a wonder now:  Frà Pandolf’s hands

Worked busily a day, and there she stands.


There are two readings to be avoided at all costs.  The first exaggerates the pentameter and emphasizes the lines as if they were end-stopped:

[That’s MY] [last DUCH] [-ess PAINT] [-ed ON] [the WALL] . . . (pause)

[look-ING] [as IF] [she WERE] [a-LIVE] [i CALL] . . . (pause)

Not only is this inexpressive, destroying both the phrasing and the emphasis (if these two things can be separated), it ignores the fact that the poet has created variations:  for example he begins the second line with [LOOK-ing], a foot called a trochee [DAH dih].  (You have already seen a trochee in [SHAD-well]).

The second disastrous approach carries the sentences past the line-ends as if they were written in prose, and it buries the rhymes:

That’s my last Duchess painted on the wall looking as if she were alive.

I call that piece a wonder now:

Frà Pandolf’s hands worked busily a day,

And there she stands.

No poem should ever be read in either of these two ways!  Ever!


One possible reading is as follows where the slant / is used to indicate a miniscule pause) and there is also a slight “dwell” on the last syllable of every line.  This “dwell” should be infinitesimally longer on [CALL] than on the other line-end syllables.

[that’s MY] [LAST DUCH] [-ess PAIN] [-ted ON] [the WALL]

[LOOK-ing] [AS IF] [she WERE] [A-LIVE]./  [I CALL]

[THAT PIECE] [a WON] [-der NOW]:/ [frà PAN] [-dolf’s HANDS]

[WORKED BUS] [-i -ly] [a DAY],/ [and THERE] [she STANDS].


The Duke of Ferrara is an arrogant man who has had his former wife killed because she didn’t cater enough to his desires.  He is showing her painting to the representative of a Count, a man who has come to arrange a marriage between the Duke and the Count’s daughter.  Is he trying to convey a warning to the Count, or is he just so self-absorbed that he is unaware of the effect of his words on others?  We don’t know, but in my view it is self-absorption:  he actually feels that he is the one who was wronged, and he is rehearsing his feelings from the past, reliving his triumph.

Being so arrogant, he always stresses “MY” and “I.”  “LAST” is stressed because that is how we say such things, highlighting their significance with our voices:  “This is the LAST time I will ask you.”  We will, as always, slightly prolong our pronunciation of the final syllable in a line, “WALL,” here.  “AS IF” is stressed on both syllables to emphasize the fact that she is dead (otherwise, the phrase could just mean that the painting, itself, is very lifelike).  “A-LIVE” bears two stresses for the same reason.  These stresses—and some of the others—are produced more by prolonging the pronunciation of the words or by slightly increasing the volume than they are by “punching” the words out.  That’s often true.  The two-stress feet [DAH DAH] such as [LAST DUCH] are called spondees, and the weakly stressed foot in the last line [-i -ly], [dih dih], is called a pyrrhic.  Any foot can be substituted for any other in a line to create variety and emphasis.  The only requirements are that the substitution fit in rhythmically, and that the regular, underlying pattern remain clear.  Varied lines shape the utterance, bringing out, or emphasizing, the meaning.  It is the mediocre poet whose lines move with metronomic regularity:  “I think that I shall never see/a poem lovely as a tree.”


We’ve looked at two-syllable feet.  Now let’s look at a three-syllable foot, the anapaest [dih dih DAH].  Because this foot is so long, anapaestic poems are usually written in the shorter tetrameter (four-foot) line.  A very familiar example is Byron’s “The Destruction of Sennacherib”:

The Assyrian came down like a wolf on the fold,

And his cohorts were gleaming in purple and gold;

And the sheen of his spears was like stars on the sea,

When the blue wave rolls nightly on deep Galilee.


We can read it like this:

[the a-SYR] [-i-an came DOWN] [like a WOLF] [on the FOLD],

[and his CO] [-HORTS were GLEAM] [-ing in PUR] [-ple and GOLD];

[and the SHEEN] [of his SPEARS] [was like STARS] [on the SEA],

[when the BLUE] [WAVE rolls NIGHT] [-ly on DEEP] [GAL-i-LEE].


Although this is beautifully handled, there is nothing very subtle about it.  Some people would insist that the first line shows elision, the dropping of one syllable of “Assyrian” (a-SYR-yun), but it is fine with four (a-SYR-i-an).  In the second line, “cohorts” takes two stresses because that is the way it is pronounced, and a stress should be placed on “wave” in the same position in the fourth line.  After all, it is the wave that is being modified by “blue,” and it is the wave that is “rolling nightly.”  It is important and needs more emphasis.  The last syllable of “Galilee” must take a stress for three reasons:  to pick up the rhyme, to mark the line’s end, and to fit into the anapaestic pattern.  (In English, you can almost always add a stress to a syllable without mispronouncing the word!)  “Gleaming,” “purple,” “gold”—these sumptuous words need to be given their full value.  The exotic image is rich with the colors of a Delacroix painting, and it is not surprising that Delacroix, a Romantic painter, chose many of his subjects from Byron, a Romantic poet.  The variation in the lines is natural and adds interest and variety without obscuring the anapaestic tetrameter foundation.


With this background, we can tackle a very difficult poem.  Let’s look at “To ____”, an extraordinarily beautiful and evocative little lyric by Shelley:

Music, when soft voices die,

Vibrates in the memory—

Odours, when sweet violets sicken,

Live, within the sense they quicken.


Rose leaves, when the rose is dead,

Are heaped for the beloved’s bed;

And so thy thoughts, when thou art gone,

Love itself shall slumber on.


It is written in two quatrains, stanzas of four lines.  The first problem—and it is a problem—is to understand what is being said.  The first stanza is easy.  It tells us that singing voices and violets both live on after they have gone—in the memory of the music and the odor.

The first couplet of the second stanza tells us that the rose leaves live after the rose, serving as a bed for the beloved.  (This invokes the idea of a lover who has piled together the rose leaves for his beloved to lie down on—or something like that.)  The second couplet of the second stanza requires us to disentangle the syntax:  “And so, Love shall slumber on thoughts of thee when thou are gone.”  The lover’s passion, the fact of loving, is personified here as “Love.”  Shelly could have chosen to be forthright, saying:  “in the same way, you will live in my loving thoughts, when you are gone.”  But instead of repeating that idea which has been carefully developed in the first three couplets, he borrows the idea of slumbering from the reference to the “beloved’s bed,” and fits the idea into it.  This enriches the poem, but it turns the final sentence on its head, making it syntactically difficult.


Now let’s turn to prosody (the study of the mechanics of versification).  A reasonable first question to ask is, “do we have movement in twos or in threes?”  And a good general rule is to look quickly at all of the lines instead of locking yourself into an effort to deal with the first one or two.  Specifically, we need to look at the two middle lines in the second stanza, for here we have straight-forward iambic tetrameter:

[are HEAPED] [for the] [be-LOV] [-ed’s BED];

[and SO] [thy THOUGHTS] [when THOU] [art GONE],


The only variation here is the weak stress I’ve indicated for “the” in the second foot—and you should know that function words such as “a,” “an,” “the,” “of,” “with,” “in,” etc. can rarely take a stress.  They are bit-part players.  (The Chinese grammarians call them “empty words” in contrast to the “full words,” the nouns, verbs, adjectives, and adverbs.)  In an actual reading of the poem, I would also place a stress on “thy” in the second of these lines.  This is because it is not her own thoughts but thoughts of her.  After all, in a slightly different form of this statement we would have [of YOU] with the accent on the pronoun.  Now let’s look at a possible reading of the poem to see how the iambic pattern fits in:

MUS [-ic, WHEN] [soft VOI] [-ces DIE],

VI [-brates IN] [the MEM] [-o -RY]—

O [-dours, WHEN] [sweet VI] [-o-lets SICK] -en

LIVE, [WITH-IN] [the SENSE] [they QUICK] -n


ROSE [leaves, WHEN] [the ROSE] [is DEAD],

[are HEAPED] [for the] [BE-LOV] [-ed’s BED];

[and SO] [THY THOUGHTS] [when THOU] [art GONE],

LOVE [IT-SELF] [shall SLUM] [-ber ON].


As we look at this, two things jump to the eye:  there are syllables both preceding and following the feet at the beginnings and the ends of some of the lines.

How usual is it to have such syllables?  Are they feet? And how does it affect our analysis?  As it happens, it is quite common to have syllables like this.  Those at the beginning of a line are usually monosyllabic feet [DAH].  (Monosyllabic feet are always regarded as being stressed.)

The first thing to note about the “extra” syllables at the end of “violets sicken” and “they quicken” is that they create feminine rhymes (two-syllable rhymes ending on unstressed syllables—{SICK-en/QUICK-en}).  When such rhymes occur at the ends of lines in an iambic context, the second syllable can just be regarded as being extrametrical, (outside of the metrical pattern).  In other words, they may be regarded as if they don’t belong to feet at all.   It is worth noting that in this poem the following foot, in both cases, is a monosyllabic foot; thus, it is as if the syllable at the end of one line completes the following iamb.  (If we decide that the lines are iambic tetrameter, beginning with monosyllabic feet and ending with a feminine rhyme we have included the notion of extrametrical syllables in that analysis.)


I have stressed “when” in lines one, three, and five because the subordinate clauses carry crucial information:  “music . . . vibrates WHEN this happens,” “odours . . . live WHEN this happens,” “rose leaves . . . are heaped WHEN this happens.”  “When” must bear a stress because it points out these important conditions.  The same sort of thing might be said of “in,” “within,” and “heaped.”  Notice that the comma that precedes “when” in each line indicates a slight pause in the middle of a foot.  This is sometimes a red flag warning us that we are misreading the line.  Here I don’t believe that that is the case.  This is just a very idiosyncratic little poem written by one of the greatest poets.  I think that he takes chances and displays unerring judgment.


Some people would claim that the “o” syllable in “violets” is elided (just as they would claim that the “i” in Assyrian is swallowed up in the Byron poem).  To do so, I think, is to reveal a tin ear.  There is one other interesting point, and that is the presence of “live” at the beginning of the last line in the first quatrain, and “love” at the beginning of the last line in the second.  These key words exhibit consonance, a term that many people use far too casually.  I generally reserve it for an identical framework of consonant sounds surrounding different vowels, “love/live,” “tried/trade,” “hurt”/heart,” ”crept/crypt,” etc.  These words tie the stanzas together, and I suspect that their effect can be felt even if their presence has not been noted.


This is a complex and beautiful poem written in two quatrains of iambic tetrameter.  A monosyllabic foot is substituted for the first iamb in six of the lines, and two of the lines end in feminine rhymes.  And that rather concisely describes what appears at first glance to be a terrible puzzle.  It is not surprising that so much nonsense has been written about this little poem.