If given a reason to develop such a process, such as attracting prey or scaring away threats or even making recognisable patterns to 'talk' to humans, what is the most likely way they would do this?
Would they be more likely to growl? Hum? Groan? Or would full on speech be possible?
Edit: Ideally it should be a repeatable sound, not one that requires the regrowth of a cell structure.
The vocalize an organism would need an air reservoir and some mean to apply pressure on it, forcing flow through an adequate aperture.
The closest I can imagine a plant can get to this is by developing some vesicles with the following features, resembling something similar available in Ecballium:
When these vesicles break open (either because an animal steps onto them or bite them) the release of gas through the opening could produce something similar to a short whistle or feeble rumble.
The noise they make is very quiet, but if you place a corn seedling in front of a laser, the sound vibrations can be detected and you can play it back through a speaker.
The sound resembles a soft crackling, and is made by the roots growing. This may not sound like vocalizing, but if the sound is played back to a seedling, it responds.
One experiment played the sound to a row of corn seedlings, and timelapse footage clearly showed that the roots were growing towards the sound.
There has been speculation on the purpose of this adaptation, and it seems to be related to direct the roots to grow somewhere, perhaps away from competitors.
Since the noise is made by the roots growing, to have audible vocalizations you would need faster growth. So, let's look at the fastest growing plant in our world, bamboo.
Scientists believe that fast growth in bamboo is caused by high coordination between cell division, cell growth and cell wall biosynthesis, which allows the plant cells to reproduce far quicker and optimizes the plants for rapid growth.
Grasses grow by elongation, but if your roots did a similar thing to bamboo, they would produce louder noises. Different plant species might produce different noises to due to differences in the density, shape etc. of the roots.
EDIT:
Ideally it should be a repeatable sound, not one that requires the regrowth of a cell structure
This could be repeatable if it had the super-fast growth I discussed earlier. The plant could have several offshoots of the stem, the cells of which would grow extremely fast. Differing speeds of growth and lengths of it would produce different sounds, and the offshoot breaks off when it is finished a single "statement". The other plant "hears" this, responds, and the cycle repeats again.
EDIT #2: Here is an excellent suggestion from Inoutguttiwutts:
Your answer becomes more viable if the plants are a hive mind. Each growth sound can form a phoneme, and each phoneme comes from a different plant in the hive. Thus individual plants communicate in monosyllabic 'grunts', but a whole plantation 'sings'
Again, an answer may lie in bamboo. Each bamboo plant is in fact interconnected to others by a subterranean system of root-like structures.
The world's heaviest organism, Pando, is a clonal colony of a single male quaking aspen, which looks like an entire forest. It uses a similar setup to bamboo.
Anyway, a plant like Pando could potentially evolve to have many "drones" which grow with extreme speed to vocalize. A single plant in the centre contains a sort of pseudo-nervous system (The system could be based on chemicals contained in the water carried by the xylem) which synchronizes the vocalizations of the drones to produce communicative messages.
Could this develop into speech? If there was a selective pressure to do so, probably. If all the plants grew this fast, there would be higher competition among species for growing space, so more complex communication might be needed. Or, if your plants have other sensory systems besides "hearing", they might be able to detect herbivores, and tell their fellow plants to secrete acid, extend spines or something.
Resonance Chamber
We know many plants can adjust the orientation of various parts in a reasonably rapid pace by adjusting the water pressure within their tissue. Sunflowers use it to follow the sun, morning glories to open the flowers when its cool, fly traps and pitcher plants to close their mouths.
Consider a species of plant that communicates by adjusting the pitch of a resonance chamber, which is sounded by the breeze. Think of how we have all blown across the top of a bottle to make it ring. If that bottle could change its shape, and thus its pitch, the plants could do so in a similar manner.
Humans would then be able to communicate with the plants by repeating the tone and duration, and facilitate communication of the plants by placing a fan near them to give them a constant breeze to communicate with.
Maybe this evolved as a means of the individual plants to monitor each other for predators, or to find the best location to grow toward for the spreading of their seed. Over time, the plants grew a means of encoding and decoding the sounds to create a sort of collective intelligence.
Travelling through a grove of such plants in a canyon might sound eerie during a thunderstorm as the plants sing to each other with the rumble of low, powerful organ music as they worship the storm that brings them precious water
The main idea here is that the plants dont necessarily have to be able to produce the wind in order to make use of audible communication. They just have to grow in a place that facilitates a reasonable expectation of breeze.
The follow-up novel to Orson Scott Card's "Ender's Game" (called "Speaker for the Dead")
has that feature in a species of trees. They use their water-transporting fibers to manipulate the frequency in which the whole tree resonates when "pinged". So another species would basically play them like a drum (very slowly), and they would change the resonance to communicate. This symbiotic lifestyle is explained by a complicated system of burial rituals, basically planting the tree in the still-living body of a dying mammal whose "spirit" will then take possession of the tree, hence the trees are the ancestors of the mammals, and that's why they communicate.
One simple solution is for them to be of a shape where the wind, or some other natural phenomenon, passes over the plant and creates noise. Barring that, the plant must be able to produce mechanical energy--movement of some kind. I immediately think of the way crickets make noise. Perhaps the plant, through expanding or contracting, may rub against itself or something nearby. This could work for tapping also. This could be caused by the plant becoming more turgid from absorbing water, or some other more complicated process.
The plant would use a violin-like vibration generator and premade (grown) resonator cavities.
Plants can move. Here is a mimosa - a "sensitive plant"
https://en.wikipedia.org/wiki/Rapid_plant_movement
If you can move you can create vibration. Envision drawing a bow across a string. A semirigid structure (like the human vocal cords, or a violin string) can vibrate.
Our speech is just modulation of vibration noise from our vocal cords using the structures of the mouth and tongue. Persons without a larynx can substitute vibration from an electrolarynx and shape these vibrations into words. Speech is stereotypically robotic. Most people have heard someone use one of these.
I have never seen nor can I find an artificial larynx played with a bow but it would work and would be a great thing for a story. I can imagine a violin-like device under the chin of the user. It can do better than the robot device because the user can play different notes or intensities according to the desired voice tone.
Ah yes - the plant. The plant would also produce vibrations like a violin. It might shape the vibrations into words as we do, but that is a lot of moving. A speaking plant might have various resonator cavities, and vibrate under the one / ones that it needed for a given vocalization. One of the earliest artificial speech generators worked this way.
http://research.spa.aalto.fi/publications/theses/lemmetty_mst/chap2.html
2.1 From Mechanical to Electrical Synthesis
The earliest efforts to produce synthetic speech were made over two hundred years ago (Flanagan 1972, Flanagan et al. 1973, Schroeder 1993). In St. Petersburg 1779 Russian Professor Christian Kratzenstein explained physiological differences between five long vowels (/a/, /e/, /i/, /o/, and /u/) and made apparatus to produce them artificially. He constructed acoustic resonators similar to the human vocal tract and activated the resonators with vibrating reeds like in music instruments. The basic structure of resonators is shown in Figure 2.1. The sound /i/ is produced by blowing into the lower pipe without a reed causing the flute-like sound.
The resonators would work as well with a vibration from a string or vocal cord or leaf. Fricatives would be hard for a plant but you could get used to speech with mostly vowel sounds.
You are asking for something completely new.
Animals have muscles. we vocalize by tightening muscles to change something that produces sound. Vocal cords vibrate at different frequences when they are stretched different amounts.
Plants don't have muscles. They can move slowly -- like mimosas can move their leaves by changing the way water moves inside them. That's slow. They can grow things in tension and suddenly release it, like a venus flytrap snapping shut on an insect. But a Venus Flytrap's trap takes hours to open if it fails to catch its prey on the first attempt.
They can sometimes grow fast, but that's wasteful as communication. Imagine you had to grow new muscle for each syllable you spoke, and the muscle provided a permanent record of your communication but was not otherwise very useful. I bet you wouldn't say a whole lot.
We haven't yet considered a way for plants to think fast. If they don't think fast, maybe they don't need to communicate fast. Of course they don't have anything like animal brains, but they are known to respond to stimuli some. If we assume plants with some kind of slow intelligence, then they can communicate slow.
So for example some leaves that curl up and expand. If it takes 10 seconds for a leaf to change state, that's a communication rate of 6 bits per minute. If a plant can control multiple leaves independently at the same time, then its maximum communication rate depends on how many it can do. If it can move 10 different leaves indepedently, then the rate goes up to one bit per second.
Plants that communicate with nearby plants could have individual rootlets that communicate with each other, and they could do it chemically. If it takes 10 seconds to send a chemical message, and it can control 10,000 independent rootlets that communicate, that's a rate of 1000 bits per second. Assuming they have a communications protocol that can interpret the signals.
Plants might be more likely to communicate with humans some way other than by vocalizing. Making audible sound costs energy and words people notice require quick changes. A plant could create a sort of vocoder if it could depend on wind to blow through it. It could manipulate vascular tissue to change the shape and make words. But if a plant takes 20 seconds to say "Hello" humans might not notice. A signal that humans can see might be more likely.
